Forums  > General  > Point72 seeds Quantopian-backed fund with up to $250mm  
     
Page 1 of 2Goto to page: [1], 2 Next
Display using:  

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-07-27 19:40
Announcement here: https://www.quantopian.com/posts/big-news-for-the-quantopian-community-managing-external-capital

I'm pretty cynical but I'm also vastly inexperienced compared to most of you on here. Would love to hear your thoughts, as I'm assuming they will be similar to the "long salmon" thread. (http://nuclearphynance.com/Show%20Post.aspx?PostIDKey=181681)

Patrik
Founding Member

Total Posts: 1333
Joined: Mar 2004
 
Posted: 2016-07-27 20:12
Reckon the key thing here is "up to". I assume from point72 side it's a cheap option, starting very small and only scales up to bigger money if it works well.

Capital Structure Demolition LLC Radiation

HitmanH


Total Posts: 423
Joined: Apr 2005
 
Posted: 2016-07-27 20:43
As you said - cheap option - but a kind of self-fulfing one - and not dependant on 'working well'. Order is:
Take a stake in ManCo / company.
Stake set
Then promise to invest up to $250m - which is a rounding error.
Watch stake in ManCo / company increase
Fund investment isn't really at risk (exit of 5pc drawdown?) - but that doesn;t touch the sides compared to the increase in ManCo investment...

Same game as seeding - returns of fund investments are normally distributed - but this lifts the right hand side, via the ManCo stake/option

jslade


Total Posts: 1070
Joined: Feb 2007
 
Posted: 2016-07-27 21:19
Pretty standard Bank backed VC play; invest in a company, claim to be a customer to add credibility. Also close to how Point72 did business when it was SAC.

A friend dragged me to a Quantopian meetup; it was lolzy. I noticed that they actually *sell* signals from zacks and such, which is hysterical as it dumps the risk (of paying for signal-worthy data) on the end user. Haven't combed through their signals enough to understand whether or not they have a decent set of regressors to build something on.

Another acquaintance; very smart guy who works in a different field, found a way to game the system more or less via volatility harvesting. No downside risk to him.

"Learning, n. The kind of ignorance distinguishing the studious."

HitmanH


Total Posts: 423
Joined: Apr 2005
 
Posted: 2016-07-27 22:05
They don't have a decent set...
And what they do is very correlated to one-another...

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-07-27 23:29
How does Quantopian even make money? I simply can't fathom any user-submitted algos being superior to what's already out there, especially factoring in tech. implementation, transaction fees, risk mgmt, etc.

Is this another SV play where it's all "free" but then all of a sudden they have stuff to sell you?

sigma


Total Posts: 105
Joined: Mar 2009
 
Posted: 2016-07-28 07:07
>> Another acquaintance; very smart guy who works in a different field, found a way to game the system more or less via volatility harvesting. No downside risk to him.

jslade,

Is it a long/short vol strategy or only short?

In case of short vol, how does he avoid downside risk?

Just a hint in case you know. Thanks

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-07-28 18:18
>Another acquaintance; very smart guy who works in a different field, found a way to game the system more or less via volatility harvesting. No downside risk to him.

Would you mind elaborating? I'm a huge novice, but I'm assuming he was able to (basically) submit a vol harvesting strategy with a smooth PNL graph / favorable risk metrics / whatever, Quantopian accepted it (maybe realizing it was better than all the more correlated strategies mentioned above, particularly as part of a portfolio), and he gets 10% of profits but isn't on the hook for losses?

HankScorpio


Total Posts: 462
Joined: Mar 2007
 
Posted: 2016-07-29 00:21
Re: Vol harvesting

perhaps you should search for the "Shannon's Demon" thread.

jslade


Total Posts: 1070
Joined: Feb 2007
 
Posted: 2016-07-29 01:04
rftx nails it. He only used price/vola information.
I think they had some kind of contest thing, and he beat everyone else by some enormous factor (most of the strategies failed completely). What I enjoyed about this is he is a sophisticated guy who knew exactly what he was doing, aka gaming the system, and was bemused at the adulation. The money is irrelevant as he has too much already; it was winning the game at something preposterously unrelated to his day job.

I've always said a lot of hedge funds are just levered S&P punts; collect the 2&20 at no downside risk until it blows up. Great work if you can find it.

"Learning, n. The kind of ignorance distinguishing the studious."

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-07-29 16:56
That's pretty funny. (It's starting to seem that to these SV types, everything is starting to look like a nail.)

But really, does anybody know how Quantopian actually makes money?

I'm really insistent on trying to piece this one together.

My thoughts:

I can't imagine it has any serious edge in the world it lives in, particularly now that some of their direct competitors can (probably) see exactly what they're doing. (Another thought: I'm currently reading the biography of Rockefeller, "Titan," and it's interesting how such an effective strategy for 150+ years has been to find your most formidable potential competition, convince them you're an ally and create a partnership of whatever form, any form will do, and keep them under your indirect watch/control so they never get a true edge. Seeing this lots with fintech as well; JPM is doing it.)

I see a few business models, not necessarily mutually exclusive.

1. When I visit a forum like HackerNews, it's clear to me there is a very large population of otherwise very intelligent and talented compsci types who think they can eke out an edge in the markets by using their brainpower. (Note: don't read the finance/markets related threads on HN unless you want an aneurism.) Quantopian basically becomes the ThinkOrSwim, eTrade, whatever, for that population. The platform that convinces generally "intelligent" retail traders that they, too, can make a lot of money, and that this new platform gives them all the tools they need. I admit the structure is a bit different this time around, but it was mentioned by jslade that they may be using this platform as a sales distribution channel. There was a BCG report out recently that I actually somewhat agreed with, even if it's obvious to some: it basically forecasts the AM industry ends up with 4 buckets of companies that survive: specialized alpha shops, beta factories, solution providers, and distribution powerhouses. Put simply - Quantopian markets itself to retail traders as a "specialized alpha shop" that you, too, can be a part of. In reality it's a bit of a combo between a solution provider and a distribution powerhouse. It aggregates lots of wannabe algo traders with cash to blow, and ultimately sells them solutions, both of their own creation and 3rd party solutions. ("But we're also helping some of you make money, see?!")

2. It genuinely does run these algo's, and in the same vein as jslade's friend: by raising money from other SV VC types who will love this type of idea, rather than from your traditional HF fundraising sources (capital introduction, personal connections, another financial company or fund, whatever), it can run outside money, gain if it works, and walk away if not. Basically, a traditional and rather unsophisticated HF that is going after an easier source of funding.

3. ???

Martinghoul


Total Posts: 852
Joined: Oct 2008
 
Posted: 2016-07-29 17:41
... 4. Where do I sign up Smiley ?

Insofar as I may be heard by anything, which may or may not care what I say, I ask, if it matters, that you be forgiven for anything you may have done or failed to do which requires forgiveness...

ronin


Total Posts: 205
Joined: May 2006
 
Posted: 2016-07-29 18:44
> But really, does anybody know how Quantopian actually makes money?

I have no clue.

But, if I had to guess, I would say on trading comissions. Everything else is marketing.

"People say nothing's impossible, but I do nothing every day" --Winnie The Pooh

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-07-29 19:21
>But, if I had to guess, I would say on trading comissions. Everything else is marketing.

The tricky thing is that, as far as I can tell, they take your algorithm and trade on it using "their" capital.

I could maybe see something where they run a tiny fund, and sell information on "retail algo's" to bigger shops (almost-but-not-quite similar to brokers paying for (or to see?) retail order flow).

I have no expertise in this area - would seeing what the current popular "retail"/"basic"/"unsophisticated" algo world thinks be informative to a shop like Point72? Is there enough volume done by these types of algo's to justify it, or is the information contained by seeing the trends in their development worth something?

ronin


Total Posts: 205
Joined: May 2006
 
Posted: 2016-07-29 21:33
Like i said, i have no clue what they actually do.

But everything about it screams marketing. Nothing about it screams prop.

Quantopian marketed successfully to Point72.

Point72 is presumably marketing to the next generation.

In the long run, the plan is presumably to have large AUM and charge for it.

"People say nothing's impossible, but I do nothing every day" --Winnie The Pooh

radikal


Total Posts: 253
Joined: Dec 2012
 
Posted: 2016-07-30 03:23
In a perfect world, you market Quantopian to outside capital so that "only the best" algos are used in the market blah blah and then you front-run that flow if it's large enough. But, napkin math on say 2.5 bps of captive flow fisting you'd need to collect doing that is 1T-ish of total flow (to recoup the 250M), which even discounted out pretty far, seems like an order of magnitude too high.

There are no surprising facts, only models that are surprised by facts

Nonius
Founding Member
Nonius Unbound
Total Posts: 12666
Joined: Mar 2004
 
Posted: 2016-07-30 07:23
how do they mitigate the (fraud) risk that some guy just reverse engineered the most amazing "backtest" that performs well on an amazingly large set of out-of-sample runs?

Chiral is Tyler Durden

sigma


Total Posts: 105
Joined: Mar 2009
 
Posted: 2016-07-30 07:52
rftx, nice summary!

>> The tricky thing is that, as far as I can tell, they take your algorithm and trade on it using "their" capital.

A potential problem can be that they will impose very tight stop losses on P&L of live algos, most likely on tick-by-tick or over a very short time-frame.

Even if an algo has an edge, it also produces volatility. For sure, the volatility of a live algo can be much higher that its volatility in the back-test.

As a result, even if you want to target a specified level of the volatility, you need to have your algo running for some time to estimate its vol.

What do you think?

deeds


Total Posts: 346
Joined: Dec 2008
 
Posted: 2016-08-03 21:51
Sounds like an update-able sliding estimate (bayesian?) is the way to go

tangentially, anybody seen a reasonable small sample volatility estimator better than the naive approach...?

sigma


Total Posts: 105
Joined: Mar 2009
 
Posted: 2016-08-04 07:53
>> tangentially, anybody seen a reasonable small sample volatility estimator better than the naive approach...?

Estimators using Open-High-Low-Close ticks are good and more robust.

Alternatively, you can apply t-distribution PDF for returns: estimate the tail parameter out of sample and use in-sample observations to estimate volatility with ML

Again, my point is that in-sample volatility and return distribution of a strategy in the back-test may not be enough for adequate estimation of volatility of the strategy in live trading. It may depend on the trading frequency of course. For sure you need to make sure that the execution of the stop-loss in the back-test is the same, or even more conservative, as for the live strategy.

deeds


Total Posts: 346
Joined: Dec 2008
 
Posted: 2016-08-04 14:44
thanks, sigma

second your point...would go further, in that, given march of time and reflexivity of markets relying simply on backtest information will not be adequate...further...could it be that fully "optimizing" with respect to back test is...sub-optimal...without taking into account online nature of real task at hand, or estimation parameters...seems shoe-horned into typical inferential statistics, made for bayes

sigma


Total Posts: 105
Joined: Mar 2009
 
Posted: 2016-08-04 15:09
Thanks, deeds

That is my concern for seeding these types of strategies with outside money: the imposed risk limits can be too tight for a strategy to survive in the long run and produce any benefits for its contributor. Also, given the fact that most likely you cannot adjust it fast without an external approval when it is live.

Does anyone have such an experience / comments?

rftx713


Total Posts: 78
Joined: May 2016
 
Posted: 2016-08-05 01:38
>As a result, even if you want to target a specified level of the volatility, you need to have your algo running for some time to estimate its vol. What do you think?

I think that makes a lot of sense (as do the additional comments)... add it to the list of why I'm so confused about how they actually make money.

From reading this thread it's clear I'm the amateur in the conversation, but I would still want to know about the size and structure of the fund overall more than anything. My logic being: who cares if it's poorly run if its existence is to serve a marketing function more than a PNL one? Or maybe, what is the fee structure on this outside capital? Are they just trying to hang on and collect fees and not blow up, while the fund also serves a marketing function for it's sales platform? I'm lost.

sigma


Total Posts: 105
Joined: Mar 2009
 
Posted: 2016-08-06 23:25
I think that the quantopian aim ultimately to become a fund-of-fund type of allocator to quant strategies. For sure they will reap benefits if they get a share of a typical 5% allocation to alternatives by large institutionals.

However, if there is a definition of a black box, this is the top of it.
As an investor in such funds, I think there are three primary type of risks:

1) Performance of individual quant strategies. Crucial is independent back-testing before putting them live.

2) Allocation procedure to best performing (in- and out-of-sample) strategies. Of course, they will aim to select 99.99% of all contributed strategies, but from the risk point of view, each strategy should not get no more than 5% of total funds. Then comes diversification among strategies.
First, they don't want to be concentrated to any particular type of strategy, say momentum.
Second, they need to stay market -neutral across aggregated positions from all strategies.
Obviously, their task is simplified if they have N uncorrelated strategies with relatively constant volatility, but I suspect it won't be the case.
Then of course it is important to account for the dynamic nature of exposures and volatilities of selected strategies

3) Operational risk. Important is:
a) How is a strategy selected to the live basket. Does it get a diversification benefit?
b) How stop-losses are triggered? Is there diversification between drawdown times?
c) How a strategy is removed once its loses its edge and how new strategy is introduced?

My biggest concerns is that these points (especially 2 and 3) cannot be really back-tested given the discretionary nature of decision making process. As a result, only after the "fund-of-strategies" is up and running live for some time we can get some understanding of its risk-reward potential.

When you ran a basket of strategies in your fund or account, you should think of them in terms of the total performance of your fund. Here, I may see the separation of interests between a strategy developer and "fund-of-strategies" manager.

Finally, there is the diminishing alpha effect. If this "fund-of-strategies" is indeed able to generate alpha, you want to be early in the game before the crowding will reduce the alpha

From Point72's standpoint it must be that they want be the first in the game in case it is indeed worthwhile. In the end, 250mm is not that big money for them to pay for staying in the first place if there will be indeed any potential allocation to these type of "fund-of-strategies" by large institutionals. If this "fund-of-strategies" will indeed prove to be a success at the initial stage, the subsequent management fees will pay off handsomely.

goldorak


Total Posts: 979
Joined: Nov 2004
 
Posted: 2016-08-07 12:37
I think there is a word to summarize all this quantopian stuff: "hype".

Strategies are nothing else than filters applied to time series. Developing filters on time series is already difficult as hell, but being able to select the right strategies (which is nothing but filtering out the filters) makes you evolve in a totally different league. I would even claim this is not the same sport anymore. You could be the best 100m runner, but still that would not make you the best 100m swimmer if you trying applying the same exact technique. Hey dude, you may think of staying above the water rather than trying running on the swimming pool's ground as fast as you can? If you want to get a chance in that different sport, you better stop thinking like institutions are thinking now. sigma is not wrong in his detailed comment, but it is just not enough. I am not going to extend too much on this topic as I consider further knowledge competitive advantage.

However, when it comes to quantopian, although they are not the worst you could imagine, they are still a number of wars behind. Their discourse sounds like the conquistadors coming to America and facing a modern end of 20th century warfare. For example, their recent paper [gold], although being a lot more advanced than whatever FoHF bs you can find out there, is just laughable at best. To add insult to injury, they actually MAKE PUBLIC criteria of interest, and of course encourage developers to concentrate in overfitting their criteria.

No, definitely, they are not up to the level needed to do what they intend doing. Hype.

If you are not living on the edge you are taking up too much space.
Previous Thread :: Next Thread 
Page 1 of 2Goto to page: [1], 2 Next