When it's normal to have no idea what your returns are

The lingua franca of asset management regardless of strategy is returns.

The lingua franca of asset management regardless of strategy is returns. But this is not necessarily true in trading. In my own 21 years of trading, it was never even brought up internally. It gets mentioned somewhere far in the background. In vague terms at best. And even then, the context is more of a hurdle than a target.

We’re going to talk about this in 2 parts.


I’ll walk through what this meant in the 3 phases of my career (with plenty of color interspersed):

  1. Being a market maker at SIG
  2. Running my own market-making group financed by a Chicago firm that backs traders
  3. As portfolio manager within a relative value volatility trading hedge fund. This last job was for an entity that was backed by LP money and therefore beholden to the concept, language, and delivery of returns. And yet, the idea of returns was alien to how I ran my business there.

Next week

I’ll connect this back to a reader question that got me thinking about all this in the first place:

How important is (il)liquidity in options when making risk-defined trades such as credit/debit spreads or buying single call/put options?


I can feel the doubt.

“Bruh, those lily pads are in different ponds, how we makin this leap?”

We gonna make it. Off we go…


My first trading job for SIG where I had my own account and P/L was an equity options market maker on the AMEX floor.

The AMEX, like the NYSE, is a specialist system. There are “posts” where the options on a list of tickers trade. The specialist is like a lead market-maker who sees the full order book and is required to make a market according to a set of rules and guidelines. The specialist designation is a bundle of obligations and privileges. They must maintain an “orderly” market, set the vol curves which then disseminate electronic bids and offers for every strike and maturity for a name to the world. They are also entitled to 40% of the volume that trades on the bid or at the ask.

Market makers stand in front of the specialist post and announce any bids or offers that improve the specialists’ market. The market-makers are allowed to participate alongside the specialist on the bids and offers if they agree that they are “on the market”. My first year trading, I was a market-maker. The major names at my post were AIG, Qwest, Eastman Kodak, Corning, Cheesecake Factory, and about 25 other stocks. One that was notably missing because it was just delisted from that post — Enron.

As a novice trader, it was really about strapping on a helmet and getting experience. I wasn’t going to be taking massive risks but in the course of trading if I saw anything noteworthy I could discuss it with my manager to see if I should be. There were 3 notable things that happened in my time there.

  1. AIG accounting scandal (this was the only time I ever talked to Jeff Yass about a trade. UBS put up a giant print in the options and Jeff dm’d me to call him with details on one of the black phones placed near the trading posts. I was scared shitless to call because I didn’t do any part of the trade and was afraid he was going to undress me for missing it. He just listened, thanked me, and hung up. I never heard another word about it).
  2. Massive buyers of Qwest teeny puts by cap structure arb flow who were trading puts vs CDS. Selling them week after week in size and overhedging the hell out of them made my year. It was also stressful because up days were painful, but the stock leaked lower and the puts barely budged.

    [A year or two a colleague ran the same playbook in Xerox in much larger size and had an amazing year. If I remember correctly, that fellow was banned from Vegas casinos for card-counting by the time he was 22. He left SIG after that year, 25 or 26 years old, co-founded a prop firm and retired very wealthy in his 30s. Our wily friend also made the news for a buying a call option on a penthouse from its owner that ended up being the highest value residential RE trade in the city where it happened. Not a small city. One funny thing I remember about people’s impression of him — he was very lazy but insanely smart. And despite my personal belief that endurance and effort > brains, there were a lot of counterexamples to this back in those days. There were savants who struck it rich and peeled off. The billionaires in the options world were the savants who worked their asses off to build businesses, but the clever cats won life. You got $50mm in 2009 and your like 33 years old. You’re gonna triple it by age 50. In flip-flops with a flip phone.]
  3. I found a dividend in Kodak that was being priced in the wrong month. Which brings you to the question — how do you pick off the people who stand next to you all day with a structure that doesn’t tip them off too quickly and also disguises that it’s you?

    [This is a separate topic but trading cultures revere cleverness. It’s a game where the goal is to take people’s money. Any harmony is just a long-run game theoretic compromise. It is fundamentally adversarial. Can you see why effective altruism starts to look like effective autism when you consider the frames traders must adopt to deal with the discomfort of their decisions? At least back then the sociopaths were more forthcoming in their intentions rather than torturing normal people with trolley problems.]

After a year at that post, I was moved to one that included Microsoft (I was there when they announced their first dividend — it was special $3 div I believe), Oracle, and Expedia. That was 2003. After 2017, 2003 was my second least favorite year. There was less action compared to prior years and markets were getting tighter (the purely electronic ISE grabbed a ton of market share). I felt deeply discouraged. I came into the business near a peak and this was my first downcycle. I extrapolated doom.

So what were my returns in those first 2 years?

I have no idea. I had a close to $1mm profits in the first year and broke even in the second year.

I don’t know how much capital I used. In fact, they wouldn’t want me to know. For example, the margin I was paying financing charges on is likely much higher than it what it was in reality considering the difference in rates a small trader gets versus a giant client like SIG. It would only make sense for the partners to effectively lend me the difference rather than allowing me the credit for funding at the firm’s rates. It’s 6 in one hand, half dozen in the other when you zoom out but it does obscure your true rate of return. It wasn’t clear how much margin you are using or how much it varied. And they had no incentive to elucidate this.

If you were looking to leave then I’m sure you could have backed into a good guess for how much capital you needed. Which is critical if you are looking to build one of these businesses since of course it’s still a capital allocation decision whose ROI must be compared to other uses for the cash. But when you are making the donuts, there’s no talk of return.

So how do you talk about the business?

It’s raw dollars. Management figures “Ok, the average MSFT market maker for SIG should trade 5,000 lots a day for a penny of edge (remember the multiplier is 100, so it’s a true $1 per lot) times 252 trading days — the spot is worth $1.25mm per year”.

[If you make less, they'll ding you in your bonus, if you make more, they say the trading was better that year so you were expected to make more — here’s your expected bonus. If you get discretionary bonuses you know the routine. You’ll get a verbal reach around in your bonus meeting, but then the number falls short of the rhetoric. I’ve been able to laugh about this for a long time now, but my wife can remember the days where my traders friends and I would plan what we would yell as we “flipped the desk over” 3 months in advance of our disappointing meetings. No matter how much you get paid, it’s always a letdown. She would joke about the Trader Wives Club where they’d have to hear us whine for 3 months before our reviews, and another week afterwards.]

In short, a trading spot occupied by a someone who knows what they’re doing has an expected p/l with a distribution. Based on the activity in the names that year, the value of the spot could be upgraded or downgraded for the following year. There’s always a dollar target and and the outcome is a debate about how much skill the trader brought to the result versus how the assumptions of volume, volatility, and competitive forces varied from the start-of-year forecast.

If the cost to man the spot makes sense compared to the expected p/l, someone will be assigned to it. Spots that are more valuable will be staffed with the more experienced/talented traders (ie you should expect that meme stocks are piloted by a prop firms’ top traders). So while there’s no concept of return, hurdles and opportunity costs are baked into the staffing decisions.

Backed by Prime

From 2008 until early 2012 I was backed by Prime International, a prop firm based in Chicago. Back then, they bankrolled about 100 traders, many in futures but there was several option pods of various sizes. I ran a mid-sized one and shared an office with their largest one (that pod was the largest market-maker in crude oil options in the late aughts).

These years were the most fun and learning I had in my career, outside the first year out of college as a trainee which was a zero-to-one combustion. I’ll save the stories for a non-print medium. I vaguely remember colleagues researching what it would take to create a small ice rink and get a penguin for the office. The fact that I would have put a 25 delta on it actually happening is a pretty good indication that this was not a normal work environment. Any job after this was gonna be a letdown but the floors’ days were numbered.

With the backer, everything was transparent. I was getting 70% of my p/l. I could see my daily margin. My shared and non-shared expenses. I could hire and fire as I wanted.

Yet again, no concept of return.

If you just looked at margin, you could back into an expectation that you should return 50-300% per year to make the gig worthwhile.

[This was in fact typical but I don’t think any pod back then was using more than $10mm and most utilizing less than $2mm. Floor trading isn’t that scalable. Ironically, the way to keep a floor well-fed is for fiduciaries to trade as if markets are more scalable than they actually are.]

Unfortunately margin isn’t a great denominator for returns. Buy a bunch of teeny options and you can mask risk. You could hunt for the cheapest thing to buy that makes a concern, defined too objectively on the back of assumptions, go away if you know what canned scenarios the risk group runs. I’m not saying this is done on a conscious or nefarious level but it’s too easy to affect the Ouija board by having a little extra preference for this strike or that maturity.

Computing return on average margin can also be weak depending on the nature of a strategy. Margin calculations are coarse proxies for risk. They aren’t custom enough for option traders. At one end they can be gamed, and on the other end they can be too conservative for arbitrage strategies (for example, no margin relief for WTI look-alike swaps if one leg is on ICE and the other on CME).

A picture is emerging. Return is difficult to compute because the denominator is murky.

We are accustomed to returns and volatilities. They let us use Sharpe ratios to measure risk-adjusted performance. But if we don’t have returns how do we get to risk limits that make sense? How do decide where to put capital?

Hedge fund days

In 2012, I moved to SF to build the commodity relative value volatility business for Parallax. It is a master fund with a host of sub strategies. LPs see returns but internally there’s no concept of returns at the strategy level. It’s just p/l that rolls up to the fund level. To be clear, this is typical for such a structure. It’s flexible.

The fund posts margin and manages risk such that the margin to equity ratio stays comfortably under 100% even in stressed conditions. Which means there’s always a cash buffer which can be held in T-bills, box spreads, or managed in any highly liquid way.

The sub-strategies margin requirements will bounce around based on the volatility and opportunities in the markets. Just to make up numbers, imagine the firm aum is $1B and runs 50% margin-to-equity. So in typical conditions, the margin requirement is $500mm.

Now consider that my margin requirement in the commodity book ranges between $20mm and $100mm depending on how much risk I’m taking. If I was a standalone fund and to be highly confident that my margin-to-equity wouldn’t exceed 80% than I’d need to raise $125mm. Most of the time I need much less, which causes a drag on returns but in the master fund structure I don’t really worry about this. The firm’s excess cash isn’t allocated to strategy as a hard constraint. The GP acts as the ultimate capital allocator internally.

So the best guess of what my returns are depend on the flawed denominator of average margin and relative to that number, they will always be worse if I’m a stand-alone fund because I need to raise far more capital than I’d typically be deploying.

How did it the business work in practice?

You’d come up with an annual expected p/l. In my case, the median was about 70% of the expectancy because I ran a positively skewed book.

[I was typically long vol convexity and often long gamma. When I was at Prime with my own money on the line, it was not uncommon to be paying 4 figure theta bills. I stood next to a guy that traded 100% of his own money that had a seven figure bill a few times a year. (Random thought but learned a lot about playing the player not the cards from him but this was a small market which he would turn into a game of heads-up no limit with the big customer. He recognized he had a lot of edge, and pushed. I hope he shifted gears when that was no longer the way you could play.)

In my fund book, you could buy a Porsche or Lambo every day with with the theta. That said — I kept a close eye on a very simple measure — don’t be long too much extrinsic premium unless there was a specific reason.

VRP language is something that feels like it comes from the asset management world not the floor trader world. There are exceptions. I knew some large short vol traders at every stop. The biggest lesson is that this game is far more artful than risk premia discourse pretends.

As for my own long vol bias and performance — this was not some trick of “Oh I lost half my years but won more in good years”. No backer, prop firm, or absolute return fund would tolerate that. I broke even or was down small 3 out of 20 years, made medium amounts most years and put up the numbers that drive the mean from the median around the GFC and the 3 year period spanning from the 2018 Volmageddon (although I wasn’t directly involved in that trade, it was the return of vol after the 2017 idiocy) through 2020. Look, the job pays when people are in pain. I don’t know what to tell you. The rest of the time I watch beta-maxxers and PE suits buy mansions. I write to feel prosocial. I don’t need trading to that feel that way. Sensing a dumb counterparty squirm, one who almost certainly was getting paid too much previously for charlatanism, is a reassuring hug from the god of markets as far as I’m concerned.

I am an agreeable human (I’m like 95th percentile on the Big Five personality test for this) but a highly disagreeable trader. Process, patience, fold, more patience, then f you sold. Boring, boring, boring, paper cuts, I hate this job, boring — violence. Put your style into simple words one day. It helps steer you back to the North Star whose light whose light you’ll most when your every decision feels like it takes you further into the dark.]

Back to the returns stuff. You peg an annual p/l target. Management implicitly considers how much risk needs to be tolerated to achieve that raw p/l, deems it satisfactory and off you go. Next year, the landscape is reviewed, growth initiatives weighed and you repeat a somewhat informal process. Any course-correction is mostly handled ad-hoc as the PM sees whether or not the environment calls for taking more or less risk. You can tell when there’s too many predators (entities who see the world in a similar way to us) vs prey (customers that have hedging or punting desires). There’s a time to hunt and time to hibernate. I say it all the time — this is a biological system not a Newtonian one.

So back to returns…it’s not quite right to think of return on margin. If you want to force a business like this into that framework you should probably just be conservative and consider how much AUM you’d need to run the strategy as a stand-alone fund. I’ll use a broad informed stroke. With a few hundred million in capital, I’d guess a strong manager with a trader mindset (as opposed to asset-manager mindset — if you’re in this business you know the difference so don’t @ me) could put up mean returns in the ballpark of 9-12% with the median between 50 and 70% of that.

[You can also see why many of these business are best housed within a fund that can flexibly allocate unutilized cash. The traders can be paid well enough to not tempt themselves into the brain damage of starting a fund of their own. To go out on your own has to be about more than money. There needs to a psychic reason to want your name on the door to endure what it takes to launch a fund.]

This is a damn good proposition because it behaves like long option position that you get paid to own. It’s unsurprising that the fees for the handful of managers who can do this (if you can even access them) are high. The proposition gets much worse if you’re taxable because it’s a short-term gain bonanza but of course many institutional allocators are tax-exempt.

Most of the risk lies in the ability for talented group of people to self-perpetuate themselves and the ability to assess that from the outside is probably almost zero. So all the usual caveats of active management apply.

Like I said earlier, next week I’ll tie this back to a seemingly unrelated question:

How important is (il)liquidity in options when making risk-defined trades such as credit/debit spreads or buying single call/put options?