Forums  > Careers  > Small team, high Sharpe  
     
Page 1 of 1
Display using:  

NeroTulip


Total Posts: 1051
Joined: May 2004
 
Posted: 2019-12-11 05:20
I've been talking to a fellow NP'er -you know who you are- and I am wondering if others are in a similar situation:

- Small team (1-10 people) at a prop firm, trading a 3 Sharpe over the last ~5 years (not selling tail risk).
- Annual profits in the $10-50 million range, no crazy HFT infra costs. We all know the guys who make $10 million with a Sharpe of 7, as long as they can use their firms' $100 million infra.
- Considering moving to a different place with lighter costs, opportunity to invest in their strategy and business, and better tax treatment.

I am wondering if there are other exemples, in HFs, prop shops, or small firms, of small teams of similar quality. Can such a small team produce these results sustainably? How do they compete with Rentech/Two Sigma etc... Why doesn't Millenium hire them all? Why don't they set up their own firm? Would they set up as a HF or a prop-trading firm?

Anyway, comments welcome, and if anyone is in a similar situation, you can always drop me a DM.

"Earth: some bacteria and basic life forms, no sign of intelligent life" (Message from a type III civilization probe sent to the solar system circa 2016)

ronin


Total Posts: 527
Joined: May 2006
 
Posted: 2019-12-13 11:27
You are asking us to evaluate a team based on three bullet points?

In all seriousness, if I had to guess - I would say they probably are selling tail risk. Just not in any form that you know about.

You said yourself they don't have some market beating infrastructure. And they are not diversifying the sh*t out of the markets, because that's what Milenium and Citadel do already, and they are not quite Sharpe 3. So what's left is tail risk.

Why doesn't Milenium hire them - presumably they looked and said "you are selling tail risk, no thanks"

"There is a SIX am?" -- Arthur

Its Grisha


Total Posts: 14
Joined: Nov 2019
 
Posted: 2019-12-13 17:38
Do the Chicago semi-automated options market making shops fall into this? I don't think they need particularly insane infrastructure on less liquid products. Just some very smart people modelling vol and some sharp traders. And while I don't know numbers first hand, I suspect the top performing desks in those firms are meeting your criteria. It also isn't uncommon for desks to spin off in the space. I guess Millenium/Two Sigma don't hire them because it's not really the same business model?

I'm also very curious on the question and am mainly speculating, so please correct my misconceptions.

sharpe_machine


Total Posts: 33
Joined: Feb 2018
 
Posted: 2019-12-13 17:56
@ronin

Is there some way to quantitatively determine how much of tail risk a strategy owns? (well, I understand it is almost a 1M$-question, but still)

longGamma


Total Posts: 17
Joined: Jan 2019
 
Posted: 2019-12-13 19:41
Ronin nailed it. They're likely short tail risk ..

@sharpe_machine
if the S&P 500 gaps down 10%, what's the strategy's p&l at open?

sharpe_machine


Total Posts: 33
Joined: Feb 2018
 
Posted: 2019-12-13 19:46
@longGamma
Are you talking about strategies with overnight exposure?

What if it is 100% intraday? (not high frequency, like ~1 trade per symbol per day)

ronin


Total Posts: 527
Joined: May 2006
 
Posted: 2019-12-13 21:04
> Is there some way to quantitatively determine how much of tail risk a strategy owns? (well, I understand it is almost a 1M$-question, but still)

Are you kidding? 1M$? Man, if it was that cheap - why, anybody would be doing it...

Not by looking at the results, no. Or if they don't want to tell you. And they don't. Which is why @nt has this conundrum, presumably.

But maybe once you start discussing things like limits etc, something may pop out as looking a bit strange. Out of place. Like, a $1mln strategy has to have $100 mln worth of orders in deep book. Or something like that.

So you have a conversation about it, and dig deeper until you are comfortable. Which may be never. It ain't easy, fundraising for something like that. And it is even harder putting money in it.

"There is a SIX am?" -- Arthur

NeroTulip


Total Posts: 1051
Joined: May 2004
 
Posted: 2019-12-15 10:14
@ronin:

To be fair, I am not asking to evaluate the team, but merely if similar teams exist in your experience. Your answer seems to be no, it's all tail risk.

I do find the situation quite exceptional and I am wondering what the catch may be. Usually the catch is tail risk, low capacity, or difficulty replicating in another firm (non competes, tech needs, etc...). It does not seem to be the case here.

"Earth: some bacteria and basic life forms, no sign of intelligent life" (Message from a type III civilization probe sent to the solar system circa 2016)

ronin


Total Posts: 527
Joined: May 2006
 
Posted: 2019-12-15 16:28
"no" is a strong word. I am not saying that.

But I woud look for other explanations. Like they say - if it sounds too good to be true, it probably is.

"There is a SIX am?" -- Arthur

gill


Total Posts: 206
Joined: Nov 2004
 
Posted: 2019-12-17 16:19
Are they liquidity takers or liquidity providers?

EspressoLover


Total Posts: 391
Joined: Jan 2015
 
Posted: 2019-12-17 18:06
On the other hand, even if it is selling tail risk, do you really care if you can put it in a thinly capitalized LLC, add a shit-ton of leverage, and constantly pull out profits? At most you can only lose a thin layer of margin capital. Imagine you put $1 in LTCM and every month redeemed capital above the initial principal. Even accounting for the blowup, you would have made pretty decent money.

At one point, if someone's good enough at disguising tail risk, then they'll probably also fool the broker/clearinghouse. At which point the tail risk mostly becomes their problem.


Good questions outrank easy answers. -Paul Samuelson

NeroTulip


Total Posts: 1051
Joined: May 2004
 
Posted: 2019-12-18 02:35
@gill: liquidity takers, although I am interested if anyone has examples that are providers too.

@EspressoLover: Yes that's possible, but not my preferred way to operate.

"Earth: some bacteria and basic life forms, no sign of intelligent life" (Message from a type III civilization probe sent to the solar system circa 2016)

gaj


Total Posts: 60
Joined: Apr 2018
 
Posted: 2019-12-18 04:35
> Do the Chicago semi-automated options market making shops fall into this?

Curious about this as well.

I know of some teams with similar profiles -- don't know their actual numbers but guessing they're in the same ballpark. My impression is their edge comes from market access. They trade in markets that are not fully electronic, so human decision making is still important here. There's also less competition than the HFT space.

ronin


Total Posts: 527
Joined: May 2006
 
Posted: 2019-12-18 15:49
> On the other hand, even if it is selling tail risk, do you really care if you can put it in a thinly capitalized LLC, add a shit-ton of leverage, and constantly pull out profits? At most you can only lose a thin layer of margin capital.

That's aggressive mate. You would be effectively defrauding the pb. I'm not at all sure they wouldn't be able to prosecute you for something like that. And that is quite apart from the fact that no pb would open a $10 mln account on its own.

> if someone's good enough at disguising tail risk

I doubt that's the case here.

@nt, I bet once you start discussing capital and limits with them, something will pop out.

"There is a SIX am?" -- Arthur

EspressoLover


Total Posts: 391
Joined: Jan 2015
 
Posted: 2019-12-19 17:05
Fraud requires either material misrepresentation of a fact or failure to disclose duty-bound facts. At least in the US, a client has no positive obligation of disclosure to a broker. Unless you make an outright statemen like "my positions are incapable of losing more than their margin capital", there is no legal basis for civil or criminal fraud. If making excessively risky bets within a corporate entity was illegal, half of Wall Street would be in jail.

If you lose more than the account value, the broker may attempt to pierce the veil and recover losses from the owner and/or parent company. But corporate personhood is extremely well-protected by Delaware and Cayman courts. As long as the entity keeps separate records and bank accounts, it's nearly impossible to pierce the veil outside criminal malfeasance. I'm not aware of a single example in the history of finance where the limited partners were personally liable for the losses of an investment fund. If this was the case, then private equity would be an unviable business model, since the funds, and ultimately the investors, could be made liable for the debts and product liability of their portfolio companies.

To the second point, if you're doing a lot of business you can definitely find some broker to give you a lot of leverage. It might not be a "prestige name", but someone's hungry enough for business to ignore the risk. For example you mentioned strategies that layer 100 levels deep in the book. It's easy to find FCMs that advertise intraday leverage 20x higher than overnight CME margin. That comes out to 500:1 leverage on ES notional exposure. And that's just for the guy off the street. If you're pushing 10k+ contracts a day in volume, most off-label FCMs are happy to give you whatever leverage you want.

But the broader point I want to make, is that even if the strategy doesn't have tail risk, you should still use this approach to quarantine the exposure of any type of black box strategy. Even without tail risk, algo trading systems can easily burn all their money due to simple operational errors. This is doubly true in the context of a third-party manager. You can layer on external risk controls, but all systems are fallible.

Good questions outrank easy answers. -Paul Samuelson

Its Grisha


Total Posts: 14
Joined: Nov 2019
 
Posted: 2019-12-20 00:58
@EspressoLover very interesting points.. Not to derail the thread but recently I got an interesting question in an interview that seems highly relevant.

I give you $1 million that you do not have to pay back if you lose it, but you keep any returns above the principal after 1 year.. there's an instrument that gives a 10% return on this in 1 year. Do you prefer this with 5% volatility or 10% volatility? What about 10000% volatility? If you have a preference for extreme vols, where is the crossover point at which they become favorable?

Nonius
Founding Member
Nonius Unbound
Total Posts: 12790
Joined: Mar 2004
 
Posted: 2019-12-20 06:05
@Nero

it looks too good to be true IMHO, but it would depend on what markets they are trading. if they are trading super liquid instruments, one way to square a sharpe of 3 with what they do is to ask for the average R^2s of their alphas, their turnover rate and the number of instruments traded (by the general Grinold Kahn stuff, but if the transaction costs are high that's not quite applicable. It would give an upper bound though).

@EL

on creditors piercing the veil and recovering losses from limited partners/owners, that's the whole purpose of structuring things properly (Bankruptcy Remote structures, SPVs and the way almost all hedge funds operate with a master fund managed by a separate asset manager/investment manager/investment adviser). Not sure that requires domiciling in Delaware or Caymans but I could be wrong.

As for cases in which investors have been able to win a compensation claim on the partners, the only cases I'm aware of are related to fraud. A bit reticent to post the details of one particular case (think a darker more evil version of Madoff), but in that case the guy was ordered to pay back $500Mn to investors. He's on the lamb somewhere in Switzerland.

Chiral is Tyler Durden

NeroTulip


Total Posts: 1051
Joined: May 2004
 
Posted: 2019-12-20 09:04
Thanks Nonius, that helps!

"Earth: some bacteria and basic life forms, no sign of intelligent life" (Message from a type III civilization probe sent to the solar system circa 2016)

ronin


Total Posts: 527
Joined: May 2006
 
Posted: 2019-12-20 16:31
> That comes out to 500:1 leverage on ES notional exposure. And that's just for the guy off the street.

Wusses. Cyprus is full of shops offering 10,000:1 leverage to the guy of the street.

It works really simple. The punter puts in $100. The "broker" let him put up a fictitious $1 mln position. The fictitious position has daily volatility $10k. It has 100% probablity of piercing through his $100 margin within hours. The "broker" pockets the $100 margin. Easiest money in the world.

But I don't think that's relevant to the topic here.

> But the broader point I want to make, is that even if the strategy doesn't have tail risk, you should still use this approach to quarantine the exposure of any type of black box strategy.

Fully agree. In my shop we run all our strategies in segregated accounts. But that doesn't mean we can happily put up a strategy with fat tails and expect the broker to eat it up when the tails blow.

"There is a SIX am?" -- Arthur
Previous Thread :: Next Thread 
Page 1 of 1