In this episode, ReSolve team members Adam Butler and Rodrigo Gordillo discuss the concept of market volatility and its implications for investment strategies. They delve into the intricacies of Carry strategies, their impact on portfolio returns, and how to navigate market volatility effectively.

Topics Discussed

• The concept of Carry in the investment world and its potential applications

• The role of Trend and Yield in creating a fund

• The importance of understanding the Carry strategy and approach

• The impact of economic risk on Carry signals and expected future returns

• Building Carry strategies using time spreads or calendar spreads

• The influence of Absolute Carry and Relative Value Carry on performance

• The effect of estimated real trade frictions on returns

• The correlation between Carry and other assets in a portfolio

• The performance of Carry in different market regimes

• The benefits of diversification and the role of Carry in enhancing portfolio returns

This episode provides valuable insights into the world of market volatility and Carry strategies. It highlights the importance of understanding these concepts for effective investment decision-making and portfolio management. A must-listen for anyone interested in deepening their understanding of market dynamics and investment strategies.

This is “ReSolve’s Riffs” – published on YouTube Friday afternoons to debate the most relevant investment topics of the day, hosted by Adam Butler, Mike Philbrick and Rodrigo Gordillo of ReSolve Global* and Richard Laterman of ReSolve Asset Management.

*ReSolve Global refers to ReSolve Asset Management SEZC (Cayman) which is registered with the Commodity Futures Trading Commission as a commodity trading advisor and commodity pool operator. This registration is administered through the National Futures Association (“NFA”). Further, ReSolve Global is a registered person with the Cayman Islands Monetary Authority.

Welcome.

2

Welcome everybody.

3

Welcome to our, latest webinar

covering Managed Futures

4

Carry: A Practitioner's Guide.

5

before we begin, we're going to give

everybody a little time to get settled in.

6

Make sure that everybody's signing

up and, I've seen the participant

7

count go up quite quickly.

8

So, it looks like we may be

ready in a couple of minutes.

9

Adam Butler: How's everybody doing today?

10

This whole webinar format is not

so good for my presentations.

11

You know, I'm struggling to

picture everybody naked over

12

Rodrigo Gordillo: the

13

Adam Butler: webinar.

14

I guess it's only you, Rod,

and that's no good for me.

15

Rodrigo Gordillo: So I wonder, I want

to make sure I can, yeah, the Q and A.

16

How's everybody doing today?

17

Feel free to get into the q and a

chat box and start testing to see

18

whether we can get your questions,

as we wait for people to come by.

19

What we really wanna do in this

webinar is Adam and, Andrew

20

wrote it very technical paper.

21

We're gonna go through a lot of stuff,

but I'm gonna be, Adam's co-pilot here.

22

So I'm gonna try to channel everybody

that's sitting in their seats right

23

now trying to think through whether.

24

You know, they might be

asking questions throughout.

25

I'll try to be that co pilot and ask

those questions, but please do answer.

26

so do ask questions.

27

I'll be reading them and I'll see if, if

I can help get that answer for you guys.

28

Adam, are you saying Rod

isn't good enough to look at?

29

Exactly right.

30

Like if he's going to look at a

naked person, you'd think at this

31

point it'd be his business partners.

32

Feel comfortable with it.

33

Adam Butler: It's just getting old.

34

I have to say that I've been looking

forward to presenting on carry for years.

35

This is my favorite, most

misunderstood strategy.

36

So it's great to get

it out in the sunlight.

37

Rodrigo Gordillo: Is it possible

to have a fund that is trend

38

and yield no stocks and bonds?

39

It certainly is possible.

40

we'll be talking less about product

today and we'll be talking about

41

more conceptually the, the carry

strategy and the carry approach.

42

Just to kind of introduce this

because I think one of the reasons we

43

were, I'm really pumped about this.

44

I know Adam's been dying to talk about

this for years is that Even three years

45

ago, nobody wanted to talk about trend.

46

It seems like that, that there's

been a big uptake on trend.

47

General audiences are

starting to understand it.

48

They see the value in it.

49

And, very few people know about carry.

50

So lots to talk about.

51

There's, I'm so excited about this for

many reasons we're going to cover, but,

52

yeah, I think there's many applications

to this concept in the investment world.

53

Adam Butler: All right.

54

We better roll because,

we've got a lot to cover.

55

Rodrigo Gordillo: All right.

56

So you want to pop up that presentation,

Adam, since you can be doing most of it.

57

Adam Butler: Okay, let's do it.

58

Rodrigo Gordillo: All right.

59

So we have a lot to cover today

and we're going to try to get us

60

through it as much as possible.

61

We are covering a white paper that has

a lot of output and a lot of back tests.

62

And so lots of this will be just kind of

reiterating the concepts through charts.

63

So even though there's a lot of slides,

we'll try to get through them quickly.

64

but if we go to the next slide,

Adam, why, you know, should you

65

be listening to me as a host?

66

I am Adam Butler's business partner.

67

Been together with him and

68

I've been in the business for almost

20 years now, and I'm president of

69

Resolve Asset Management, Global.

70

Been a-co contributor to a lot of

white papers and podcasts and blog

71

posts on Resolve and, you know,

writing articles as well now on the

72

return stack, website and, and so on.

73

So been managing quantitative

investment strategies for many years.

74

And today I'm going to help facilitate

the conversation with the co authors.

75

If we go to the next slide, Adam.

76

So for those who haven't read the

white paper, I would encourage

77

you to click on the link.

78

Maybe Ani can push that through for

you guys to be able to click on that.

79

If it's not, if you haven't already,

it'll be a good thing to have side by

80

side as we go through this presentation.

81

and the white paper was

written by Adam Butler.

82

Our CIO here at Resolve Global and Dr.

83

Andrew Butler, our resident PhD in

Resolve Asset Management Canada.

84

yes, they are related.

85

Genius does run in the genes.

86

So our two propeller heads did

a fantastic job at creating a

87

comprehensive review and framework for

different ways of looking at a carry.

88

And everything that we're going

to present in this presentation,

89

all the performance stats will

be sourced from this white paper.

90

So just, click on that.

91

There's also data that you can download.

92

So if you can kind of double check

everything that we review here.

93

Adam Butler: We should also mention

there's an advisor, summary of this

94

on the InvestResolve blog, which

is a little bit more accessible.

95

We've cut out some of the more, nebulous

sections of the white paper and made

96

it just grounded a little bit more.

97

So, feel free to check that out too.

98

Rodrigo Gordillo: and Ani, you can put

the link to that in the chat as well.

99

If we could next slide, Adam, for me,

you know, as always with, any sort of

100

work that you do with the investment

in the investment universe, Adam,

101

the next slide, I still don't see it.

102

we have to be cognizant of, all

of the important disclaimers.

103

We're going to be presenting

a lot of hypothetical results.

104

This is merely a research project that

tries to shine a light on a premium that

105

may add value to people's portfolios.

106

There's many ways to skin this cat

by no means it is a, an offer to,

107

buy a fund or anything like that.

108

So really it's about

covering that research.

109

Next slide for me, Adam.

110

So please do read those disclaimers

and recognize that there is risks in

111

everything that we do in this space.

112

And then I'll tell you quickly

what we're going to tell you.

113

We're going to talk about what

is basic carry, how we define the

114

three basic definitions of carry.

115

How we can practically implement all

these carry strategies in a portfolio and

116

incorporating with other asset classes.

117

And then from then on, we'll really go

through some analytical framing, you

118

know, how to put these things together

to make them work in real life with

119

real trading, and then from then on,

it'll be back test showing performance

120

analysis scenario and regime analysis.

121

So a couple of case studies on

the return stacking side and then

122

conclude with some questions.

123

Okay, so next slide for me, Adam.

124

Before we do begin, let's

start with a little poll.

125

Ani, would you mind

pushing through that poll?

126

I want the audience to, I'm just trying

to get a gauge as to how many people

127

had heard about Diversified Carry

before downloading our white paper or

128

being in this, webinar or the invite.

129

And also if you can answer the second

question, which is if you had heard

130

a carry, how many of you do you

currently invest in a carry strategy?

131

Give it a couple of minutes there.

132

I've never done a poll, so I'm just

gonna let that linger there for a bit.

133

And maybe we can come back to it, Adam.

134

but, Ani, you let me

know when it goes, when

135

Adam Butler: Yeah, I think Ani closes it

out and then it shows you the results.

136

So, it's probably long enough.

137

It's only two.

138

Rodrigo Gordillo: Yeah, Ani, let's go.

139

Two quick questions.

140

Let's see if we can find the answers.

141

Okay, what do we have?

142

So, have you heard about

diversified care before?

143

Yes.

144

So most people, about two

thirds have said, yes.

145

One third said, no, that's a good ratio.

146

that's probably above average.

147

We have an above average

intelligent crowd here today.

148

And you currently invest in diversified

carry strategies only around 20%.

149

So that's kind of what we would expect.

150

hopefully, you know, we certainly do and

have been talking about carry strategies

151

and have those in our back pocket.

152

And if you go to our website, you

can find some, but, more and more

153

solutions will be coming out very soon.

154

All right, so this is good to know.

155

Let's hand it over to

our lead author here.

156

Adam, why don't you take it

away and tell us what Carry is.

157

Adam Butler: Okay.

158

What is Carry?

159

Yeah, so Rodrigo is going to step in

and ask questions and seek clarity

160

if I kind of missed something.

161

But let's start with basic definition.

162

Carry is what you expect to return on

an asset if the price doesn't change.

163

In the white paper, we use an

apartment building investment to

164

illustrate the concept, right?

165

Because you buy an apartment building,

typically you're going to own it

166

for many years, maybe decades.

167

Maybe you're not so concerned

with what the price will be when

168

you sell it in 20 or 30 years.

169

In the meantime, it is generating

a lot of cash flow for you, the

170

rents on the apartments, right?

171

Well, going a little bit more into public

markets, equities are expected to deliver

172

cash flows in the form of dividends.

173

Even equities that don't currently

pay dividends are priced on the

174

basis that eventually they're going

to return cash flow to investors.

175

Bonds.

176

Almost all of them pay a coupon.

177

That is the carry on bonds.

178

And we're going to demonstrate that

in the concept of futures markets,

179

commodities, equities, currencies, bond

futures, all of these can be expected

180

to pay or, absorb carry at, at different

points in time for different reasons.

181

So let's get into an example.

182

All we've done here is.

183

Sort of illustrated a

simple futures market.

184

One futures market that has a,

let's say it's copper, right?

185

Copper is currently trading at 3

dollars and 50 cents in a spot market.

186

So if you're going to go buy.

187

A ton of copper, you're going

to pay 350 a pound, say, right.

188

And then we've got a future on copper and

you, someone wants to buy the future on

189

copper, to take delivery in September.

190

And that's the point that's sort

of out on the right here, right?

191

So here's the spot price and

here's the futures price.

192

And, they're paying a little bit less.

193

In the future, then the current spot price

for dynamics that currently exist in the,

194

in this market, which we'll get into.

195

And then over time, if the price of copper

doesn't move, if it stays where it is,

196

we would expect the futures price to rise

to eventually hit the price of copper.

197

And therefore this market where the

price of the future in the future is

198

lower than the spot price or the near

term futures market contract, we expect

199

that to have positive carry as the price

converges to spot over time, right?

200

So in this case, we're sort of saying

we're going to earn a little over

201

3 dollars as this distant future

becomes less distant over time in

202

terms of time as time rolls forward.

203

But futures markets can take many shapes.

204

So you'll, you'd imagine copper futures.

205

You might be able to buy copper for

delivery and maybe, you know, it's

206

May now, maybe you can buy copper

for delivery in June, in September

207

and December, in June of 2026.

208

And so you can plot the price of each

of these different futures markets.

209

On a chart going out through time.

210

And that describes the term

structure of that futures market.

211

So in this case, you know, the

contracts going out into the future are

212

priced at successively lower prices.

213

And as you go out sort of far enough,

they begin to rise a little bit.

214

And that's just very typical futures

markets tend to have a curve in them,

215

reflecting the supply demand dynamics of

the market at different points in time.

216

Rodrigo Gordillo: Right.

217

And so this has got kind of just

thinking about putting the two charts

218

that we just went through together here.

219

There's many opportunities to

measure carry, assess carry to

220

define what carry is for this

particular fictitious contract, right?

221

And you're going to get, you're

going to get a reading on that.

222

And carry is really about.

223

You're making allocation decisions

based purely on that yield, that futures

224

yield, much like if you think about in,

trying to select securities based on

225

some sort of a shareholder yield, right?

226

So I know that these are done very well.

227

Meb Faber has done a lot

of work on this, right?

228

You're just measuring the type of

yield that the company as a whole is

229

pushing out and selecting those assets.

230

Solely on that, not on price momentum,

not on value, just purely on that.

231

And while much like, shareholder

yield provides positive expectancy

232

above cash, we find the same

thing when we select based purely

233

on this type of futures yield.

234

Adam Butler: Right.

235

And in this case, this futures curve is

what we might call backward dated, where

236

the few, the prices in the future are

lower than the nearer term or spot price.

237

And again, with this future market, we

would expect this to deliver positive

238

carry as these distant futures begin

to converge on the higher spot price

239

over time, all things equal, right?

240

Of course, it never works

out exactly like that.

241

The price is, The near term

contract and spot change over time.

242

The prices of the distance contracts

change over time, but on average, we

243

expect this general drift to occur.

244

Just like on average, we expect

equities to have a positive drift,

245

but they go up and down over time.

246

Rodrigo Gordillo: And two things on that.

247

Number one, the, on this shape of the

curve in that case, We're backward

248

aided, we're going long those contracts.

249

If it's flipped on the opposite

side where it's upward sloping,

250

it would be the opposite.

251

We'd be looking to short

those contracts, right?

252

Cause those will, they will

gravitate towards zero.

253

And yes, just to round off what you

said, shareholder yield, finding a stock

254

that has a strong shareholder yield

does not guarantee that the price plus

255

shareholder yield is going to make you a

positive return at the end of the year.

256

It just tends to be that way over time

257

Adam Butler: on average.

258

carry, it's sort of lived outside the

Overton window for several decades.

259

There, you know, there were lots of

managed futures funds that did either

260

indirectly or directly use carry as

a signal to inform their portfolios.

261

And in fact, when we survey the offering

documents for a variety of funds in

262

the SOC gen, CTA index, We find that

carry is mentioned second only to trend

263

in terms of the frequency that it's

mentioned as a signal that it informs

264

the, the trades that they make and the

portfolios that they hold over time.

265

So, you know, it's not as esoteric

as, as many people believe.

266

And it also, in a managed futures

context, is, Considerably different

267

than what many old timers might remember

as being the sort of idea of currency

268

carry, which for a while was the

idea of, shorting the currencies in

269

low yielding, regions, like the U.S.

270

dollar or the Japanese yen and buying

emerging market currencies that typically

271

have, Higher local interest rates, right?

272

So that's not the carry that

we're that we are implementing

273

currency carry in broadly that way.

274

But when you expand a canvas to

include equities, bonds, and a wide

275

variety of different commodities,

the carry strategy takes on a very

276

different profile as you'll see.

277

So what drives carry?

278

We talked about equities and

bonds, you know, in equities,

279

it's the dividend yield.

280

That is reflected in the futures

term structure for bonds.

281

It's the, we're only dealing with

government bonds in this context.

282

So it's the, you know, whether the

bond cash treasury term structure

283

or the guilt term structure.

284

or the bund term structure

in Europe, for example, has a

285

positively sloping yield curve.

286

So if the 10 year yield is higher than

the three month treasury, for example,

287

that would, we would sort of consider

that to be a positive yield curve, and

288

it would have positive carry in bonds.

289

At the moment, you know, we're

in a bit of a strange situation

290

where near term yields Are actually

higher than most longer term yields.

291

So bonds currently are typically

measured to have negative carry.

292

but again, that's rather unusual

over the past 30, 40 years.

293

In commodities, there's a convenience

yield, which is sort of the convenience

294

that the speculators offer to producers

in order to take on the price risk.

295

So producers can sell their production

forward, have some certainty about

296

the price that they're going to get

for that production, and they can

297

go and raise capital for to invest

in new projects, that sort of thing.

298

And in currencies, it's just

the difference between the.

299

the short term interest rate in the

jurisdiction you're borrowing in.

300

In our case, we're only using U.

301

S.

302

dollar crosses.

303

We're always borrowing in the U.

304

S.

305

dollar to invest in a foreign

currency or borrowing in a

306

foreign currency to invest in U.

307

S.

308

dollars, depending on which of

those has a higher interest rate.

309

Now, I'm not going to dwell on this.

310

This is just why do equities and bonds

need to have a positive long term carry?

311

Well, because in order to invest in

equities and bonds, You need to move from

312

very liquid cash into illiquid securities.

313

You don't know what you're going

to be able to sell those securities

314

at some point in the future.

315

If you need liquidity quickly, you may

have to take a hit on the price you

316

would realize for those securities.

317

There's also inflation uncertainty.

318

We don't know what inflation will

be like in the future and you need

319

to be compensated for locking your

money up for a long period of time.

320

And also you also, you need

to defer consumption, right?

321

Instead of buying something that you

want today, you're deferring consumption.

322

Putting it in savings vehicles or in

investments, hoping that those investments

323

will grow over time, but you need

to defer what you want to buy today.

324

So that's standard stocks and bonds.

325

Commodities, just to dig in a little

further, it's very accretive for commodity

326

producers to sell their production

forward often many years into the future.

327

Imagine, a mining company wants

to develop a new copper mine.

328

And.

329

By the time you get the environmental

permitting, all of the engineering spec'd

330

out, you do all the assays, et cetera,

to figure out what kind of mine you want

331

to build, it's probably 10 or 15 years

before you get to first production.

332

So the copper companies will sell a

good portion of the expected production

333

from those mines forward in order to

block in the economics or a substantial

334

portion of the economics on that project.

335

And they have many projects going

on in many different regions for

336

many different metals, et cetera.

337

Energy companies are doing the same thing.

338

Grain producers are doing the same thing.

339

And this is highly accretive because

it lowers the, variability of

340

their earnings over time and gives

investors certainty and that higher

341

certainty for investors lowers the

cost of capital to the producer.

342

They're able to go to the debt market and

raise money at lower interest rates and

343

go to the equity market and raise money at

higher multiples, lower cost of capital,

344

higher ROI to the producer over time.

345

So it's kind of a win situation.

346

Speculators are providing, they're

insuring the producers against

347

price fluctuations and they earn

a premium on this insurance.

348

And that's why we expect this over time.

349

Now, the commodity premium can be

positive or negative depending on

350

the short term dynamics in a market.

351

We saw energy prices go negative in

352

when there were all these pipelines

leading to storage facilities and the

353

storage facilities were totally full

and they would, were selling oil at

354

negative prices in order to make room

for new oil coming from pipelines

355

that had to go somewhere, right?

356

So there's short term supply demand

and dynamics that can make the term

357

structure for commodities positive or

negative And make it more attractive

358

to be short than long or vice versa

and that's why in carry strategies,

359

sometimes you want to be long a market

if it has positive expected carry short

360

of market, if it has negative expected

carry, if you're short of market that has

361

negative expected carry, you're expecting

to earn a positive return on that.

362

Rodrigo Gordillo: Yeah.

363

So I'll just kind of tie

this all up in a bow here.

364

I think generally speaking, we

think that carry signals do provide

365

insight into future expected

returns because they should be

366

compensated sources of economic risk.

367

So, while these things may

get crowded at times, right?

368

If too many people go into this one trade

in a particular market, the yen, U.S.

369

trade or whatever, it's not rational

economically for a risk premium

370

like carry to get arbitraged away.

371

Because As Adam kind of alluded

to here, there are players, there

372

are willing participants here that

are getting economic benefit for

373

hedging their risk and the other

side for taking that risk, right?

374

So it would require for this not to

be a risk premia, it would require

375

parties to be, willing to bear risk

with zero expected compensation, which

376

is not how the economy works, right?

377

So I think that's a good basis

for this whole carry thing.

378

Adam Butler: Yeah, so carry is

more of a classical risk premium,

379

then it's a lot harder to make

that same sort of case for trend.

380

There are different reasons

why we think trend exists.

381

Carry it's a little bit clearer

that this is a risk based premium.

382

So, for the purpose of our

experimentation, we use this diverse

383

universe of different global equity

markets, major global government

384

bond indices, a variety of major

currencies and, commodities in

385

the energy and metal sectors.

386

We didn't include any, grains or

softs or other more, out there,

387

commodity sectors for, you know,

liquidity reasons, et cetera.

388

In practice, there are commodities within

other sectors that are plenty liquid.

389

That, you know, could be used in

scalable carry strategies, but

390

this was our experimental universe.

391

Rodrigo Gordillo: And just as we go

into a lot of the analytics here,

392

I think this is a good question

somebody's asked for us to clear up.

393

So the question asked is it considered

a carry trade when, for example,

394

commodity, a commodity pool owns

treasuries as a form of collateral

395

for the futures contract, right?

396

So this is, what he's alluding to

is that if you X ray a fully funded

397

futures strategy or commodity strategy,

it is anywhere between 80 and 90%

398

Treasuries, short term treasuries

that you're earning yield on.

399

And then the remaining cash is

used as collateral to buy and

400

sell those futures contracts.

401

And so I guess I'll get my 2 cents

out Adam and then maybe you can

402

correct me, but it's important to

note that what we're going to be

403

presenting here is excess returns.

404

So we are not including the carry on

cash that would exist if we were to port

405

this strategy over to a traditional fund.

406

So this is, am I right in saying

this is excess returns, Adam?

407

And so what excess returns

mean is the returns above cash.

408

it's what you would get, if you

were to just run the strategy

409

without any cash yield whatsoever.

410

So it's not the, to answer your

question, this is the yield you get

411

on the treasuries in a pool is not

considered carry in the traditional

412

sense, in the sense that we're

going to be talking about here.

413

Now, can we make a case for

it being carry if you're fully

414

funding a fund that uses carry?

415

Yeah, I guess you could say that

it is a carry, but it has no

416

specific, quantitative strategy.

417

It's just, it just happens to be

along for the ride rather than it

418

being an explicit bet on carry.

419

Adam Butler: Yeah, that's

a really good point.

420

So how do you build carry strategies?

421

There's actually a few

different ways to do it.

422

One way is using time spreads or

calendar spreads where you want to be

423

totally neutral exposure to a market.

424

And you will just, for example, go

long the near term contract and short

425

the far contract, and then just take

the ride of the short term contract.

426

the longer term contract converging to

the shorter term contract over time.

427

That's one, way to do it.

428

Another is cross sectional carry.

429

We're going to go through that in quite

a bit of detail, which is typically

430

implemented at the sector level.

431

So for example, if you've got eight

different equity markets, you're running

432

a cross sectional equity carry strategy.

433

You're going to be long for equity markets

and short for equity markets all the time

434

to main, maintain that sector neutrality.

435

And same for long and short

energy markets, long and short

436

bond markets, et cetera, right?

437

So that's a cross sectional

or sector neutral strategy.

438

And then the third way is a time

series carry strategy, which is the

439

one that we're gonna spend the most

time on here today, where you're

440

allowing the portfolio to get, you

know, a little bit more crowded.

441

In the, on the long side, if most of

the markets in a sector or most of the

442

markets in the portfolio have positive

expected carry and you're allowing, you

443

know, get more short in on a net basis.

444

If more of those markets, have

negative expected carry than

445

positive expected carry, right?

446

So again, time series, strategies

allow for sector exposure to drift

447

higher or lower into negative territory

over time in response to how the

448

underlying markets are, you know,

expressing positive or negative carry.

449

Whereas a sector neutral strategy.

450

All the markets in the sector could

have positive expected carry, but you're

451

still enforcing the constraint that

half of them need to be held short in

452

order to eliminate any sector exposure.

453

Okay.

454

And we'll see how that impacts

strategy performance over time.

455

So within these, cross sectional

or time series strategies, we also

456

divide it up in terms of measuring

carry on an absolute basis.

457

Which is what we've been talking about

so far is the term structure of the

458

futures for a market positive sloping,

implying negative carry or negative

459

sloping, implying positive carry.

460

so that's sort of absolute carry

and have a more traditional way

461

of people thinking about it, but

there's also a relative value carry.

462

So if we look at the very

long term average of the term

463

structure, so for example, gold

is usually in, has a positively

464

sloping term structure, slightly

positively sloping term structure.

465

So the carry for gold in an

absolute sense, usually implies

466

negative, slight negative carry.

467

But if instead you look at the long

term average term structure and you

468

measure carry relative to that long

term average, then when it's above

469

average carry, then you'd go long.

470

When it's below average

carry, you'd go short.

471

and so it's just a slightly

different cut on this concept.

472

Then there's how you want to

transform that carry signal, right?

473

So a really simple way would say,

well, I want to be one unit long.

474

If the, or one volatility unit long,

as we'll talk about a little bit

475

later, if carry is positive, one

volatility unit short, if its carry

476

is negative, that's binary signals.

477

Or we could be long or short in proportion

to the, just the strength of the raw carry

478

measure, or that the degree to which carry

is above or below its long term average.

479

And then also we can rank the

markets in, in the portfolio by

480

their carry as well and use the

rank score, positive or negative

481

rank score as their, carry measure.

482

So there's all these different sort

of ways to skin the cat and they all

483

deliver, they all sort of capture the

same underlying phenomenon, but they do

484

things from slightly different angles.

485

And therefore they provide some

diversification benefits when

486

you combine them all together in

a portfolio, as we'll discuss.

487

So just to reinforce this

concept of absolute carry.

488

This is, again, just kind of

what we've already been talking

489

about most of the time so far.

490

These two futures markets, these term

structures, imply negative carry in

491

both cases, but the blue line has

a higher negative carry or expected

492

negative carry than the black line

in this case on an absolute basis.

493

So this is like raw or absolute carry.

494

In contrast,

495

Rodrigo Gordillo: Can you

go back for a second, Adam?

496

Adam Butler: Sure.

497

Rodrigo Gordillo: So what you're

saying here is there's two contracts.

498

So in a carry portfolio, that's trying to

decide whether to invest based on carry.

499

All things equal.

500

If they have the same volatility, same

type of correlation, the blue line at the

501

top would receive a higher weighting than

the black line, but they both receive a

502

negative would be shorting both of them

because they, they are, they, you know,

503

they're going to go from the high price

in the future and roll down to spot.

504

Is that fair?

505

Adam Butler: Yeah, that's a good point.

506

And so if it were a binary, transform,

then actually they would both be short

507

and have the same volatility adjusted

weight in the portfolio, right?

508

if they were ranked or if we were using

the raw score, then the market represented

509

by the blue line would be expected to

have a higher, negative weight in the

510

portfolio than the black line, right?

511

For the reasons it's discussed.

512

So that's good.

513

Now this is relative value carry.

514

We call it carry Z because

we're actually doing a Z score.

515

And a Z score is just what is the

current value relative to the mean value.

516

So that difference divided by the

amount that value varies over time or

517

the volatility of that value over time.

518

Okay.

519

So that's why we sort of use

carry Z score and relative value

520

carry kind of interchangeably.

521

Rodrigo Gordillo: Let me clear

that up a little bit more.

522

What do you mean by mean value?

523

So basically the, where is the carry

today relative to a historical average?

524

Adam Butler: Exactly.

525

Yeah.

526

So we just take the historical average

of the entire term structure, and then

527

we look at, well, is it, is the carry

higher or lower than what it, what the

528

carry typically is over time, right?

529

So in this case, the blue line represents,

higher than average negative carry.

530

The green line, while it also indicates

negative carry, it's lower than

531

negative, average negative carry.

532

And therefore we would be short the

blue line and long the green line in

533

this case, because we're now measuring

carry relative to its long term average

534

black line, not on an absolute basis.

535

Okay.

536

Okay.

537

So now the fun part, we're going to start

with an examination of whether, carry has

538

existed in each individual market sector.

539

So equities, bonds, and

different commodity sectors

540

and currencies on their own.

541

And then we're going to kind of begin

to combine things together and see

542

how that everything kind of works

together in a stepwise fashion.

543

So this will reinforce, because

we're dealing with markets in sectors

544

that have very different ambient

risk, like obviously bonds are

545

going to have very different long

term average volatility than say

546

natural gas or crude oil or copper.

547

we're going to, we're going to

scale all of the markets in each

548

portfolio to have the same volatility.

549

And then we're also going to scale

the volatility of each portfolio

550

to have the same target of 10

percent annualized volatility.

551

And when we scale them every

single day, we're evaluating the

552

volatility of the underlying markets.

553

They're correlation to one another.

554

And then we're using those estimates

to scale that portfolio to, the

555

target that best approximates

what we've estimated from the

556

portfolio in the very recent past.

557

Rodrigo Gordillo: Right.

558

So in that sense, you're not letting

the maniacs take over the asylum, right?

559

You're making sure that your bets

are equalized across the sector.

560

And then you also don't want to have any

asylum be too big over another asylum.

561

And you're getting that kind

of like you're equalizing

562

the risk for the assignments.

563

Adam Butler: So that all the

sectors have the same volatility.

564

Exactly.

565

But this is why, even though

we're scaling to target 10 percent

566

volatility in these portfolios, it

never actually gets to exactly 10%.

567

Because there's some error in our

estimation of what the portfolio

568

volatility is going to be in the next

period, every time we estimate it.

569

Right.

570

So we get close.

571

But we rarely get dead on, right?

572

And then we're just scaling, use a fairly

near term estimate of volatility based

573

on what happened over the last 40 days.

574

And we're using, because it's an

exponentially weighted moving average,

575

the nearer term returns matter a bit more

in our estimate than the returns that are

576

more distantly in the future, in the past.

577

Rodrigo Gordillo: And

you talk about the floor.

578

Adam Butler: Oh, yeah.

579

The floor just means that there are

some times when, you know, a market

580

seems to have just extraordinarily low

volatility over a short period of time.

581

And when you're deep in that low

volatility tail, that's often a

582

sign that you're misjudging the

true volatility of the market.

583

So we never let our estimate of volatility

go below the fifth percentile of our

584

measured volatility over the full

period of history prior to that date.

585

Rodrigo Gordillo: So that you're

not levering up a single security

586

that has never been, never shown or

exhibited that low level correlation,

587

or not never, but very rarely.

588

Adam Butler: Okay.

589

So again, these are just sort of

starting on, sector neutral or

590

cross sectional start strategies and

strategies that are using absolute

591

or raw carry to measure carry, right?

592

So raw carrying currencies and all the

way up through raw carrying equities.

593

Now you can see that enforcing

sector neutrality has a penalty.

594

We do not want to take on any sector risk.

595

And, you know, that's okay across most

of the sectors historically, but a few

596

of the sectors have either very low or

slightly negative historical returns

597

if you're not willing to take on some

sector risk when the, the carry skews

598

towards long or skews towards short

in any given sector over time, right?

599

they still all do relatively well

with the exception of equities.

600

but not quite as well as we'll see when

we adjust the measure of carry for,

601

you know, against the long term average

carry for each market in the sector.

602

We're still holding the sectors market

neutral in this case, but we're sorting

603

them based on the degree to which

they're, the carry is above or below the

604

long term average within each sector.

605

And this seems to have, deliver better

performance over the very long term.

606

Now on a time series sense.

607

Now we're allowing the sector exposure

to drift positive or drift negative.

608

If there's a preponderance of markets

that have positive carry or negative

609

carry at any given moment, right?

610

We do that using raw carry as our

signal, then the performance generally

611

improves across the board and we see

the same phenomenon with relative value

612

or carry Z, signals where allowing

that sector exposure to drift higher

613

or lower is long-term accretive.

614

As we're now accepting more sector risk

as well, when we aggregate up all of

615

the sectors together and we just look at

the cross-sectional carry for equal risk

616

weighting all of the different sectors

and are using a cross-sectional, portfolio

617

approach, then you can see that the carry

Z outperforms the, the regular carry,

618

which is consistent with what we kind

of saw at the individual sector level.

619

And when you combine the regular carry and

the relative value carry Z signals into an

620

ensemble for sector neutral strategies, it

actually rolls up pretty well with a long

621

term Sharpe ratio of around 0.55, right?

622

But still it doesn't compare to the

performance of time series strategies

623

where we're now allowing sector

exposure to drift over time, both

624

raw carry and carry Z strategies.

625

Both perform very well

on a time series basis.

626

When you roll them up together,

they do even better with a

627

sharp ratio in the range of 0.

628

9.

629

And even when you combine the sector

neutral strategies with the time

630

series strategies, because of the low

correlation, you still preserve the

631

majority of the performance you get

from the pure time series strategies.

632

when you combine everything together,

just the power of diversity and

633

ensembling, which we're going to

discuss as we go forward, we pretty

634

well use ensembles in all of our

strategies everywhere for this reason.

635

It should be noted by the way, because

we brought up the idea of you know,

636

raw signals, binary signals, and rank

signals, that in each of these cases,

637

we're just combining strategies based

on raw, strategies based on binary,

638

and strategies based on rank all

together within individual sectors,

639

within individual cross sectional

versus time series ensembles.

640

And within the total

ensemble portfolio, right?

641

So this is gives you a general idea of

the performance of all of these time

642

series and cross sectional portfolios.

643

over time you see the, you know, the

cross, one of these cross sectional

644

portfolios, just didn't perform very well.

645

Again, if you're not willing to accept

sector risk, then you're removing

646

what turns out to be a meaningful

component of the carry signal over time.

647

So that was individual sectors.

648

Most managers also sort of contemplate

or actually focus their portfolio

649

management on the total portfolio

and not just on individual sectors.

650

So in this case, we're allowing,

let's say, you know, all of the.

651

energy markets had negative carry, but

all of the bond markets had positive

652

carry or some mix, whatever we're allowing

those markets that have positive carry,

653

to be held in positive weight in those

markets with negative carry to be held

654

in negative weight in the portfolio

without sector constraints, right?

655

It's just, you're going to be held in

the direction of your raw carry or your

656

relative value carry in the portfolio.

657

We're going to show an inverse volatility

weighted version of this, which is just

658

using, instead of just applying inverse

volatility weighting at the sector level.

659

Now we're going to apply this

at the total portfolio level.

660

So we're measuring the volatility

of every market in the portfolio.

661

The exposure is going to be

the carry score, or the carry Z

662

score, divided by that market's

volatility estimate at that moment.

663

And then we're going to scale the,

Exposure of all markets in the

664

portfolio to hit a target of 10 percent

annualized volatility based on our

665

best estimates at any given moment.

666

Okay.

667

So that's naive or inverse

volatility weighting.

668

We're also going to explore a mean

variance optimization of the same concept.

669

In that case, we're trying to maximize

the total amount of both long carry

670

exposure and negative carry exposure.

671

So short exposures while simultaneously

trying to minimize the total

672

portfolio volatility, right?

673

The way that you would in a

typical mean variance optimization.

674

So it's just a carry score

is our return estimate.

675

And we're just trying to minimize

the total portfolio volatility while

676

maximizing our exposure to carry.

677

So the naive weight inverse

volatility, and then an optimization

678

we're going to get to after that.

679

So starting with the

inverse volatility waiting.

680

So now we're only exploring

time series versions of this.

681

We're going to leave the sectoral

sector neutral versions of these behind.

682

We're going to accept the fact

that we're going to have some

683

Sector exposure over time.

684

If all of the markets in a given or

most of the markets in a given sector

685

have positive carry, we're going to

have positive exposure to that sector.

686

And we're just going to live with that.

687

And we're going to also harvest

the excess premium we've got

688

from accepting that risk.

689

So looking at.

690

normal carry, but now examining just using

the carry score as a continuous signal,

691

or binary carry, is it positive or is

it negative, and then rank carry, so the

692

score becomes the rank, then you can see

all of these do quite well, you know,

693

in this case, the raw carry doesn't do

as well as the rank carry, the binary's

694

somewhere in the middle, we're going to

see that this is random noise, and that

695

all of these Carry signals and transforms

are basically, they all give you

696

approximately the same strength of signal.

697

It's just, some of them have done

worse or better recently or what

698

have you, but it's just noise.

699

So exploring the same thing for relative

value carry, we see, you know, a

700

different, well, we're preserving the

order here in terms of sharp ratio, but

701

the relative value carry in this case

tends to do a little better than the,

702

raw carry in some cases, a little worse

in other cases, again, emphasizing that

703

we are using all of these different cuts

at the same phenomenon, not because any

704

of them have better or worse expectancy

over time, but because they all do

705

slightly different things at different

times and diversify one another, but

706

they all do well as you can sort of

see from their long term profile.

707

Now transitioning to mean

variance optimized portfolios.

708

We're again considering both volatility

and correlation to minimize portfolio risk

709

while maximizing total portfolio carry.

710

Just using raw carry

with various transforms.

711

All do very well, and they tend to

do a little bit better than the naive

712

portfolios that are not accounting

for correlation differences over time.

713

Carry Z showing equally strong performance

with mean variance optimization.

714

And you can see these are a little bit

more tightly grouped over time, then the

715

inverse volatility or naive weighting,

again, they do slightly better than

716

naive portfolio weightings, but for the

most part, it's just noise and what is

717

better is to combine a naive weighting,

which makes fewer assumptions about

718

the portfolio with the mean variance

optimization, which makes a couple of

719

more assumptions about the portfolio.

720

Mainly that we, that correlation

estimates are a little bit persistent.

721

So when we're measuring the current

correlations, those correlations

722

are likely to be approximately

the same in the next period.

723

Combining those different approaches

is highly creative, right?

724

For diversification reasons.

725

And when we do combine them, we

observe smoother performance, higher

726

sharp ratios over time, right?

727

So we're just, this is just building up

again to our final meta, meta ensemble,

728

but just focusing on all of the different

inverse volatility portfolios, right?

729

using regular carry, all of them

using relative value carry, Z carry.

730

And when you combine them

both, you go from about a 0.

731

85 sharp to a one sharp.

732

This is just with the inverse

volatility, but ensembling.

733

Right.

734

And there's the green line above

both of the, constituent ensembles.

735

Rodrigo, I don't know if that's confusing.

736

Yeah.

737

Rodrigo Gordillo: No.

738

So I just want to, I don't think it's

confusing, but I think for those who

739

haven't read our research in the past.

740

I think it's important to put a

stake on the ground right here and

741

make sure that everybody understands

the difference between over

742

optimizing and narrowly data mining.

743

And creating a robust portfolio

that is likely to work out of

744

sample, meaning in real life.

745

And it may seem like all these

layers that Adam's been talking

746

about is about, you know, getting

more narrow and more optimized, but

747

in fact, it's the complete opposite.

748

It is the idea of trying to be

broadly correct about capturing

749

the carry signal rather than being

specifically wrong and assessing

750

all of these individual parameters.

751

Finding the one that had the best back

test and choosing that one, right?

752

So an example that I often use with regard

to the value of being humble about your

753

ability to capture any signal, the way

we're doing here through ensembles is, I'm

754

sure everybody a couple of years ago heard

that the, that we captured as humanity,

755

the first image of a black hole and the

headline said the event horizon telescope

756

captures the first image of a black hole.

757

And it was a very well defined black hole.

758

It's exactly what we imagined high def.

759

It was a beautiful image.

760

What few people know is

that wasn't one telescope.

761

The event horizon telescopes is actually

hundreds of telescopes across the world

762

in many different sites that are capturing

different types of signals, radio

763

waves, infrared, you know, they're all

measuring the black hole in their own way.

764

And what the team had to do is

over a couple of months in the U.S.

765

is grab all of that data.

766

Put it together,

eliminate the error terms.

767

And what they got was that beautiful

image that we received in the news.

768

Any single telescopes image,

if you watch the, there's a

769

documentary on Netflix is garbage.

770

It's doesn't make a lot of sense.

771

It's kind of the outline, right?

772

So it really is ensembles is the most

robust way that we have found as humanity

773

in terms of noise to ratio, noise to

signal, to make sure that we are broadly

774

correct about what we're trying to do.

775

Adam Butler: Yeah, no, that's

a really great metaphor.

776

and I love it that you saw

the opportunity to use that.

777

And this is just the most widely

document phenomenon in, data

778

science and machine learning.

779

So the winners of Kaggle competitions

invariably every time are

780

using ensemble type techniques.

781

All right.

782

It's just vastly superior to view a

problem from a wide variety of different

783

angles and aggregate the signals up when

you aggregate the signals, you reduce

784

the noise and emphasize the signal.

785

And so that's all we're doing

here with at the portfolio level.

786

Great point.

787

So now it's just the optimization.

788

It's also, I think, worth saying

that when we're portfolio optimizing,

789

we do this every single day, right?

790

So we're, you know, it's today.

791

We look back over the last Several days,

several months, we're estimating the,

792

variances and correlations of between

all the different markets at that time.

793

We're estimating the carry and

we're forming a new portfolio.

794

Then we move forward one step, we

look back again, and we, you know,

795

we're, we use that information

to form a brand new portfolio.

796

So we're constantly rebalancing into

a portfolio seeking to emphasize

797

or maximize carry while minimizing

portfolio volatility, right?

798

So when we're doing that, you know, when

we ensemble all the different transforms

799

of raw carry, We do very well, carries the

relative value that does well when you put

800

them together, you go from kind of a 0.

801

9 sharp ratio to almost a 1.

802

1 sharp ratio.

803

Historically gain just the power

of approaching it with ensembles

804

and the ensemble line above

either of the constituents line

805

scaled to the same volatility.

806

Now, you know, an actual question

to ask at this point is, well, you

807

don't know if you've been reading the

disclaimers, but they're all showing

808

gross, returns gross of estimated

trading costs and commissions.

809

So, you know, it's a good question.

810

Well, do these returns survive?

811

Estimated real trade frictions, real trade

slippage, commissions paid, et cetera.

812

And, so, you know, we've been

813

So we've got seven or eight years of

live data from our own trading that

814

we are able to use to get really good

estimates on the cost of trading these.

815

And then there are papers that

we lean on for the cost of.

816

You know, trade frictions on

different markets going back

817

to earlier points in history.

818

And we're able to net these out

and get estimated net results.

819

It's also important to know that we

are, we do with some smoothing, right?

820

So when we ensemble all of these

different, approaches together, they

821

all recommend slightly different

Portfolios at any given time.

822

And what that means is it averages

out the amount of trading that

823

you need to do from day to day.

824

So that alone sort of reduces the amount

of trading and therefore the amount of

825

trade friction that you experience, the

amount of commission you pay, and that

826

demands you place on the market to absorb

the liquidity that you're sourcing.

827

We also smooth the waste through

time using a 5 day exponentially

828

weighted moving average.

829

We find that smoothing like this has no

effect on performance, but does have a

830

nice effect on reducing trading frictions.

831

so when we apply these smoothing

and ensembling techniques and we

832

also embed our, trade slippage

estimates, then we see that we lose

833

about 1 percent a year in terms of

returns, which works out to about 0.

834

1 sharp ratio.

835

and everything kind of just drops by

1%, you know, slightly larger drawdown,

836

slightly lower sharp ratio, et cetera.

837

But I mean, this is just a tremendously

resilient strategy once you back

838

out estimating trading costs.

839

Right.

840

so a common question is great.

841

But if carry is very highly correlated to

the other assets I hold in my portfolio,

842

it may not be very useful still.

843

Right.

844

So it's important to wonder

how the correlation experience

845

for carry evolves over time.

846

Here we plot the,

rolling one year or late.

847

Oh, sorry.

848

Rolling three year correlation between

the carry strategy and the S& P 500.

849

And the U S 10 year treasury future.

850

And you can see that, you know, it

does go through multi year periods

851

where bonds have a positively sloping

yield curve and our carry bond

852

exposure is predominantly positive.

853

equities, the dividend yield on

most global equity markets is higher

854

than their local short term rates.

855

And so they have positive carry.

856

And so it, you know, we have

a proper ponderance of long

857

equity exposure or vice versa.

858

Right.

859

So it does fluctuate over time.

860

That said the long term average

correlation between stocks and

861

bonds and carry is about zero.

862

Yeah, sorry,

863

the correlation between carry and

trend is in the neighborhood of 0.

864

3 to 0.

865

4, depending on the, the frequency

that you're measuring at.

866

Right?

867

So, whether you measure daily returns

or monthly returns, et cetera,

868

it's in the neighborhood of 0.

869

3 to 0.

870

4, which is still very much in

the range of a strategy that is.

871

Where two strategies can be combined and

be nicely complementary to one another,

872

which we'll see a little bit later on.

873

Now, it's important to examine how Carry

performs in different market regimes.

874

And we define regimes in a few

different ways, but a common way Is

875

inflation currently trending higher

than expected or lower than expected?

876

And is growth currently coming

in a little higher than expected

877

or a little lower than expected?

878

And therefore we can divide things broadly

in this kind of four different regimes.

879

And I think, Rodrigo, you had a poll

question that you wanted to ask.

880

Rodrigo Gordillo: Yeah, Ani, if you don't

mind pushing the next poll question,

881

why don't you go to the next slide?

882

the question is, Given everything

that we've reviewed, where do you

883

think, carry loses money, right?

884

Cause everything, everything that this

chart shows is really, you know, we

885

can expect gold and commodities to do

well in rising inflation environments,

886

but we can likely expect them to lose

money generally in lower inflation

887

environments and low growth environments.

888

And so there's winners and losers.

889

Even when you think about trend following,

you know, we can be very clear about.

890

When there are very persistent

trends, almost all the time when

891

there's a bear market, you can count

on trend to, to likely be there

892

and provide really strong offset.

893

So there's an intuition there.

894

I'm just curious to know from the crowd,

what the intuition is for carry here.

895

What is the, what regime

does it lose money in?

896

All right, given that we

have Not a lot of time, Ani.

897

Why don't you push that through?

898

Okay.

899

All right.

900

So that's interesting.

901

Okay.

902

So we got pretty evenly distributed.

903

Why don't we show, what we

actually found there, Adam?

904

Adam Butler: Yeah.

905

So in fact, we generated the equity line

for the carry strategy, conditioning

906

on each of these different regimes.

907

And so each of these lines represents

the cumulative growth of the carry

908

strategy only during periods that are

aligned with, you know, inflationary

909

growth, deflationary growth, inflationary

stagnation, or Deflationary stagnation.

910

So the flat periods here are when the

strategy is not in that regime, right?

911

And then it moves up or down based

on how it performs conditioned

912

on being in that regime.

913

And what we see is that in broad strokes,

carry is not really very sensitive

914

to any of these, these broad regimes.

915

and that's partly due to the fact

that Because of the diversity

916

of markets, That are held in the

portfolio and the propensity for

917

the portfolio to be relatively

diversified most of the time, it ends

918

up having a very stable return stream.

919

We don't have a lot of really big,

monthly or quarterly losses or gains.

920

And so that the historical frequency

distribution, if you kind of compare

921

it to trend or to stocks, is a little

bit more normal or Gaussian in shape,

922

which is kind of the holy grail of

what you, I mean, people would maybe

923

prefer positive skew and I can get

behind that, but you know, just avoiding

924

negative skew, we think is a big win.

925

And when you, again, when you

combine carry with trend, with

926

equities, with bonds, Then that

distribution becomes even more normal.

927

it's also, you know, curious,

how does carry perform during the

928

best and worst period for stocks?

929

So what we've done here is sorted the,

returns on the S& P 500 into their

930

worst quintiles on their left, going

all the way up to their best 20 percent

931

of, orders on their, on the right.

932

Okay.

933

And you can see, obviously the light

blue line is the S& P 500 in its worst

934

quarters, it does the worst, right?

935

But turns out carry and trend do just

fine during the worst quarters for,

936

or have historically done just fine.

937

Interestingly, in the second

worst quarter, neither carry nor

938

trend really does much, right?

939

In, then in the sort of, top three

quintiles, not quarter, quintiles, The

940

carry and trend both have a tendency

to do reasonably well and obviously

941

these are very good quarters for stocks.

942

For bonds we see a similar profile

generally sort of agnostic to the how

943

bonds are doing in any given quarter.

944

Both trend and carry tend to do relatively

well even in the worst bond quarters

945

and then they go on to do actually

quite well in the best bond quarters.

946

And then you wanted to

add to those, Rodrigo, or

947

Rodrigo Gordillo: no, just kind

of broad, broadly speaking,

948

we're kind of on time here.

949

So,

950

Adam Butler: yeah, let

me sort of zip through.

951

We just wanted to go through

the profile of these strategies

952

during the worst drawdown periods

for both equities and bonds.

953

So this is the, October,

954

So the S and P 500 global financial

crisis drawdown, and you can see

955

carry It was kind of like going

sideways for the early part and then

956

went on to deliver nice returns.

957

during the tech rec, Carry did very well.

958

Very nice offset.

959

COVID crash was particularly

challenging for Carry.

960

It was probably the most challenging

period for Carry, as the, authorities

961

both on the fiscal side and on the

central bank side were way behind

962

in terms of implementing policy

to keep up with the news flow.

963

And, what we find is that carry strategies

are a little bit more susceptible

964

to miscommunication or blunders.

965

by central banks, right?

966

So, WEN has carried on particularly

badly when central banks have either

967

been behind the curve or they've been

misreading the messaging from the markets.

968

And in this case, obviously, the

authorities were very behind the

969

curve during the COVID crash, and then

they moved extremely aggressively,

970

probably more than the market

expected, immediately after the crash

971

and and things went in a different

direction than the market expected

972

and that wasn't very good for carry.

973

Eventually evened out and went

on to deliver very solid gains.

974

For treasuries, This is just

the post COVID bond bear market.

975

so actually that wasn't too bad for Carry.

976

Recovering from the

global financial crisis.

977

So Carry did fairly well during the

global financial crisis, but once

978

the global authorities stepped in,

they implemented quantitative easing,

979

markets began to settle, then Carry

kind of went sideways, struggled

980

for a little bit before recovering.

981

This is the one to focus

on, the bond massacre of 94.

982

Because this is a, an example of where

the Fed was miscommunicating or not

983

communicating with the market about their

intentions and about their expectations.

984

And they came in with a ver with

a surprise rate rise at aggressive

985

surprise rate rise, caught the

bond market off guard, and.

986

You know, bonds basically

crashed overnight and

987

everyone was sort of offside.

988

and that was an example of where carry

kind of struggled in the short term before

989

again, going on, and doing very well.

990

we'd be, I think, leaving people,

wondering if we also didn't examine how

991

carry worked alongside trend and alongside

equities of bonds in a stacking framework.

992

So, you know, because carry and trend

both have low correlation to both

993

stocks and bonds, They're just both

really accretive when stacked on stocks.

994

And here you see both carry stacked

on stocks, trend stacked on stocks,

995

and a 50 50 combination of carry

and trend stacked on stocks.

996

Obviously, all three looking

very attractive historically.

997

Same story stacking on top of

bonds, just very attractive.

998

And, turning up a negative return over

the recent bond bear market period

999

into reasonable positive returns.

01:04:09,029 --> 01:04:14,369

Just isolating the performance of

carry and trend stacks on equities,

01:04:14,849 --> 01:04:21,499

obviously boosting equity returns,

lowering equity risk, or not

01:04:21,859 --> 01:04:27,819

boosting equity risk by a meaningful

amount, despite the higher returns.

01:04:28,674 --> 01:04:32,954

And trend and carry combining to be a

little better than either on its own.

01:04:34,044 --> 01:04:40,724

Similar with, with bonds and just

combining everything together.

01:04:40,724 --> 01:04:47,244

50 50 stocks, bonds, 50 50 carry

trend, just has a, an astonishingly

01:04:47,964 --> 01:04:50,169

attractive historical profile.

01:04:50,949 --> 01:04:51,089

Rodrigo Gordillo: Yeah.

01:04:51,089 --> 01:04:54,069

And what's important here when we

think about stacking is a lot of people

01:04:54,079 --> 01:04:55,719

think, okay, I'm stacking returns.

01:04:55,719 --> 01:04:57,199

I'm also stacking a lot of risk.

01:04:57,229 --> 01:05:01,419

But I think what we need to point

out here is how little, extra

01:05:01,419 --> 01:05:06,439

risk is taken to stack a hundred

percent of these factors, right?

01:05:06,439 --> 01:05:14,029

So in the first column there 18.09,

S& P 500 plus, you know, 10 percent

01:05:14,029 --> 01:05:16,399

volatility targeted carry is at 20.

01:05:16,399 --> 01:05:17,529

66.

01:05:17,538 --> 01:05:23,709

So not a lot more risk is taken to double

01:05:23,709 --> 01:05:28,549

With all the caveats that, you know, it

can at any given time correlate and so on.

01:05:28,549 --> 01:05:32,619

But it is the benefits of diversification,

the zigging and the zagging.

01:05:33,179 --> 01:05:33,359

Right?

01:05:33,359 --> 01:05:38,299

Two asset classes and or strategies

that make positive, have positive

01:05:38,299 --> 01:05:41,829

outcomes, have an expected positive

return, but move differently from

01:05:41,829 --> 01:05:46,699

each other to create lower, low

volatility, high return, strategies.

01:05:46,719 --> 01:05:51,422

So examine that table, examine it

in the white paper and, and then

01:05:51,422 --> 01:05:54,824

reach out for questions if you want

to get more granular than that..

01:05:54,839 --> 01:05:57,205

Adam Butler: Yeah, I think it's

worth adding quickly that, just

01:05:57,205 --> 01:05:59,773

think of a carry on equities, right?

01:05:59,773 --> 01:06:01,323

So carry versus trend on equities.

01:06:01,903 --> 01:06:04,443

So you think about an equity

bull market, equities are rising.

01:06:04,963 --> 01:06:09,423

As they rise over time, they're

getting further away from what, you

01:06:09,423 --> 01:06:15,183

know, levels where trend would flip

from being long to short, right?

01:06:15,723 --> 01:06:18,713

Carry is a little different

as equities rise and rise

01:06:18,713 --> 01:06:21,673

toward a peak in a bull market.

01:06:22,298 --> 01:06:27,668

The equity dividend yield

is getting lower and lower.

01:06:28,348 --> 01:06:34,098

And oftentimes as we're coming into,

a peak in equities, it's corresponding

01:06:34,098 --> 01:06:35,908

with the Fed raising rates.

01:06:35,908 --> 01:06:37,428

The economy is getting too hot.

01:06:37,798 --> 01:06:41,818

The Fed is raising rates so that

the yield on cash is rising.

01:06:41,818 --> 01:06:46,758

Well, then the yield on stocks is

declining at some point, the yield on cash

01:06:46,758 --> 01:06:50,293

is exceeds the dividend yield on equities.

01:06:50,633 --> 01:06:56,483

So while trend continues to buy into

the equity rally, there's a point

01:06:56,493 --> 01:07:02,283

at which Carry comes in and starts

getting short equity markets as the

01:07:02,283 --> 01:07:04,803

cash return exceeds the dividend yield.

01:07:05,163 --> 01:07:10,093

So it ends up being at least mechanically

having the potential to be a nice offset

01:07:10,493 --> 01:07:12,563

for what's going on the trend side.

01:07:12,573 --> 01:07:16,163

That's just one example how,

you know, carry can mechanically

01:07:16,163 --> 01:07:19,333

diversify trend in equity markets.

01:07:19,333 --> 01:07:21,913

And there are different types of

examples in different sectors.

01:07:22,573 --> 01:07:24,253

Rodrigo Gordillo: And look,

another question that just kind

01:07:24,253 --> 01:07:28,163

of falls across the same vein

here, which is roughly speaking.

01:07:28,163 --> 01:07:32,913

The question is about carry, trend,

gold, you know, should gold be

01:07:32,943 --> 01:07:34,613

avoidable if you can allocate to carry.

01:07:34,663 --> 01:07:40,368

and, you know, our view has always been,

these are all idiosyncratic risks that

01:07:40,388 --> 01:07:43,648

you should probably add to your portfolio

because we don't know the answer to that.

01:07:43,648 --> 01:07:48,208

And in fact, I remember vividly a recent

gold and was it two or three years

01:07:48,208 --> 01:07:54,398

ago when gold was rallying and, you

know, trend strategies were long gold.

01:07:54,628 --> 01:07:56,118

And carry was shorted.

01:07:56,148 --> 01:07:59,468

Carry was wrong because the

carry was negative for gold.

01:07:59,478 --> 01:08:04,078

So no, I don't think necessarily if

gold is going, is doing its thing,

01:08:04,158 --> 01:08:08,188

there will be times when carry is dead

wrong on that strategy, I promise you.

01:08:08,608 --> 01:08:12,338

And so the idea of just eliminating,

if you believe that these, that

01:08:12,378 --> 01:08:15,748

assets that you can stack on top

are one of two things, are have a

01:08:15,748 --> 01:08:18,268

positive risk premia and are lowly

correlated to everything else.

01:08:19,198 --> 01:08:21,327

Any one of those two

will, will be a benefit.

01:08:21,898 --> 01:08:27,818

If you expect gold to be, you know,

zero returning real returns, but you

01:08:27,818 --> 01:08:30,667

can stack it on top and it happens to

be non correlated to everything else.

01:08:30,728 --> 01:08:32,138

It is accretive to the portfolio.

01:08:32,238 --> 01:08:34,837

So I think the answer is always yes and.

01:08:35,388 --> 01:08:38,308

Adam Butler: And the average

long term correlation of gold

01:08:38,358 --> 01:08:40,268

to stocks and bonds is zero.

01:08:41,298 --> 01:08:44,778

The average correlation of gold

to carry and trend is zero.

01:08:45,848 --> 01:08:49,747

It just protects against 'em, it's

the only thing that can protect

01:08:49,747 --> 01:08:51,528

against a certain kind of risk.

01:08:52,228 --> 01:08:54,508

I think gold belongs in every portfolio.

01:08:55,038 --> 01:09:01,608

And, you know, obviously talk to your

advisor, but to me, yeah, it's highly

01:09:01,608 --> 01:09:06,098

complimentary along with all of these

other diversification opportunities.

01:09:07,098 --> 01:09:09,688

Rodrigo Gordillo: Why don't we wrap

it up, Adam, and then we'll see if we

01:09:09,688 --> 01:09:11,218

have, you know, you go ahead with the

01:09:11,218 --> 01:09:13,667

Adam Butler: benefit of carry

strategies in the portfolio.

01:09:13,667 --> 01:09:14,988

and yeah, we can take questions.

01:09:15,247 --> 01:09:15,788

Rodrigo Gordillo: Yeah, obviously.

01:09:15,788 --> 01:09:17,448

Look, we're just to reiterate, right.

01:09:17,448 --> 01:09:20,428

In conclusion, I think carry has

a unique place in the portfolio.

01:09:20,428 --> 01:09:26,688

It is an under loved strategy that, I

think many people have tried to bring

01:09:26,688 --> 01:09:31,087

to market and failed, and we're trying

to like, do our best to really present a

01:09:31,087 --> 01:09:33,587

thoughtful case for why it is so unique.

01:09:33,598 --> 01:09:34,577

Why it's so useful.

01:09:34,798 --> 01:09:37,468

If you're at, people ask me all the

time, you have your bonds, you have

01:09:37,478 --> 01:09:40,577

your equities, you have your trend,

what's the next thing you would do?

01:09:40,988 --> 01:09:42,288

it's always been carry.

01:09:42,348 --> 01:09:46,148

And there's enough, video footage of

Adam just pounding the table on this

01:09:46,148 --> 01:09:47,707

over the years that, you know, it's true.

01:09:47,707 --> 01:09:51,417

And we've been at, we have used

it for, for a long time right now.

01:09:51,417 --> 01:09:55,238

So now the question is, you know,

is this webinar is one of the things

01:09:55,238 --> 01:09:58,708

that, will give us a good reading as

to whether there is an appetite for it.

01:09:58,738 --> 01:09:59,448

I hope there is.

01:09:59,998 --> 01:10:03,478

and if you design it properly,

you design it thoughtfully, you

01:10:03,488 --> 01:10:07,978

do ensembles, then it can be just

as a creative, not caveat emptor.

01:10:08,038 --> 01:10:09,748

It is normally distributed roughly.

01:10:10,088 --> 01:10:11,478

It has volatility.

01:10:11,528 --> 01:10:13,088

It will have drawdowns, right?

01:10:13,088 --> 01:10:16,548

So if you do a 10 vol, Let's

say it's a sharp of one.

01:10:16,778 --> 01:10:20,348

A good heuristic here is to

say, okay, so the return is 10%.

01:10:20,378 --> 01:10:23,678

What is a three standard deviation

event for strategy at 10 vol?

01:10:23,678 --> 01:10:25,978

It could be like 10 minus

10 is zero minus 10.

01:10:26,448 --> 01:10:29,327

Two standard deviation is negative

10 minus 10 is negative 20.

01:10:29,388 --> 01:10:32,128

You know, something a bit higher than

that is probably a good heuristic as

01:10:32,128 --> 01:10:33,768

to what to expect in terms of drawdown.

01:10:34,318 --> 01:10:38,508

One hopes it doesn't happen at the same

time as what you're matching it up with.

01:10:38,618 --> 01:10:39,018

Okay.

01:10:39,018 --> 01:10:40,338

So there is risk involved.

01:10:40,338 --> 01:10:46,988

This is not a, Panacea, it is a unique

diversifier to add to other many things.

01:10:47,878 --> 01:10:50,198

So, just recap.

01:10:50,198 --> 01:10:51,408

Look, we do a lot of this stuff.

01:10:51,458 --> 01:10:52,968

we tend to go long form.

01:10:52,988 --> 01:10:54,278

We're 15 minutes over the webinar.

01:10:54,298 --> 01:10:55,858

That's not surprising to me at all.

01:10:56,188 --> 01:10:58,408

It's a miracle we got here in

such a short amount of time.

01:10:58,408 --> 01:10:58,608

I thought it

01:10:58,608 --> 01:10:59,508

Adam Butler: was a 90 minute webinar.

01:10:59,908 --> 01:11:01,138

I thought we were doing so well.

01:11:01,398 --> 01:11:03,818

Rodrigo Gordillo: No, sadly that's,

that's incorrect, but people are here.

01:11:03,818 --> 01:11:04,468

So that's good.

01:11:04,468 --> 01:11:05,428

If you want to learn more.

01:11:05,723 --> 01:11:07,113

Go to our website, investresolve.

01:11:07,133 --> 01:11:07,583

com.

01:11:07,923 --> 01:11:09,163

we have just revamped it.

01:11:09,163 --> 01:11:11,663

So there's a lot of research

for you guys to dig into.

01:11:11,673 --> 01:11:13,303

The white paper is available there.

01:11:13,303 --> 01:11:15,293

The executive summary is available there.

01:11:15,653 --> 01:11:18,313

There's a couple of videos of

Adam kind of like in two minutes

01:11:18,313 --> 01:11:22,508

describing the benefits of risk

parity versus carry, you know, there's

01:11:22,508 --> 01:11:23,848

some interesting dynamics there.

01:11:23,848 --> 01:11:25,158

I've answered a few questions.

01:11:25,558 --> 01:11:29,428

I typed out as Adam was talking a few

questions about risk parity and carry.

01:11:29,438 --> 01:11:32,238

If anybody wants to take a look at

those, we have our book available on

01:11:32,238 --> 01:11:37,608

Amazon, and then you can explore our

strategies that, that span far and wide

01:11:37,698 --> 01:11:42,077

across, you know, evolution strategies,

which is, All encompassing, long, short

01:11:42,077 --> 01:11:43,998

market, neutral managed future strategy.

01:11:43,998 --> 01:11:45,268

You have a carry program.

01:11:45,708 --> 01:11:49,128

You have more kind of all terrain

strategies, all types of stacking stuff.

01:11:49,128 --> 01:11:51,938

So take a look at our

strategies page, explore that.

01:11:52,358 --> 01:11:54,318

if you have any questions,

you can reach out to the team.

01:11:54,758 --> 01:11:58,478

and, yeah, I'll see if I can,

the last couple of seconds here.

01:11:58,478 --> 01:11:59,888

There's a few more questions, Adam.

01:12:00,178 --> 01:12:00,508

is there.

01:12:00,948 --> 01:12:04,178

Is there any risk that keeps carry?

01:12:04,178 --> 01:12:05,048

I'm going to combine two.

01:12:05,048 --> 01:12:05,448

Okay.

01:12:05,528 --> 01:12:08,628

Earlier on in the presentation,

there's a question about, and I think

01:12:08,628 --> 01:12:09,818

this is super important to address.

01:12:10,128 --> 01:12:14,788

are you worried about the, short

volatility character of carry with that's

01:12:14,998 --> 01:12:17,498

he assumed, I think this was early on

before you went through everything.

01:12:18,068 --> 01:12:20,868

and then this other question is what

is the risk that keeps you up at night?

01:12:21,178 --> 01:12:21,448

Right.

01:12:21,448 --> 01:12:23,518

So let's address both of those.

01:12:24,648 --> 01:12:24,818

Adam Butler: Yeah.

01:12:24,818 --> 01:12:25,238

I mean,

01:12:25,548 --> 01:12:27,678

Rodrigo Gordillo: does carry

have a short volatility tilt?

01:12:28,368 --> 01:12:30,598

Adam Butler: Carry, yeah.

01:12:30,598 --> 01:12:32,678

Shortfall carry.

01:12:33,288 --> 01:12:39,827

If you run carry on individual sectors

like currencies or equities or bonds, then

01:12:39,827 --> 01:12:44,268

you do see some left tail events for sure.

01:12:44,378 --> 01:12:46,768

And there's good reasons why they occur.

01:12:47,098 --> 01:12:50,128

because there's a flight to

quality during financial crises.

01:12:51,118 --> 01:12:51,678

the magic.

01:12:51,678 --> 01:12:55,818

in these carry strategies is

the diversity of the holdings.

01:12:56,778 --> 01:13:02,368

It's that when there's a crisis in

equities, often there's an offsetting

01:13:02,628 --> 01:13:09,228

move in bonds, or there's an

offsetting move in gold or energies.

01:13:09,673 --> 01:13:12,383

Or metals or what have you.

01:13:12,663 --> 01:13:17,943

And, as a result of that, you actually

do observe a, quite a normal, return

01:13:17,943 --> 01:13:26,073

distribution on carry in stark contrast

to the, I think, boogeyman version

01:13:26,263 --> 01:13:31,663

that those who had heard of carry a

few decades ago had in their mind.

01:13:32,033 --> 01:13:32,503

and.

01:13:33,833 --> 01:13:34,843

What keeps me up at night?

01:13:36,163 --> 01:13:39,633

I think what keeps me up at night

is that people would lean too

01:13:39,633 --> 01:13:43,433

heavily into any one strategy.

01:13:43,673 --> 01:13:50,603

You know what the miracle here

is the ability to combine stocks

01:13:50,743 --> 01:13:56,733

and bonds and, you know, maybe

gold with carry and trend and.