For those of you who aren’t aware of the types of funds. When someone says CTA, they’re usually a momentum fund or capture some well known futures effects. A lot of these guys are a strange form of levered beta and not really alpha because a lot of their methods work best…
1/n
when lots of others use them so strangely enough not the most advanced or accurate model is the best model because really it’s just this greater fool theory flow. You can pretty easily frontrun this shit, because they typically ease into positions because they have…
2/n
Quite large books, and they compound their positions based on the simple rule of add to winners and cut losers (basically the entire premise behind momentum where losing money is a get out signal and making money is the get in signal). Anyways, if you know there is…
3/n
Lots of CTA flow there and it’s making money, just make the momentum trade and quickly cut losses. It’s something that there’s not a literature on but when should you cut losses, assuming losses are a sign of no more alpha? It’s a problem of building a really good signal…
4/n
Estimation (Kalman? CNN?, savgol?, wavelet?, I won’t give it away, since this sort of problem comes up a lot where finding a suitable signal estimation is key, and I don’t want to give away all my alpha). But generally the goal is to use ML to find the ideal exit point…
5/n
IMO this is a nice merger between algoquant and trading flow for a cyborg trading operation. CTAs say they’re in commodities as the name suggests but really this type of fund will be in everything from treasuries to futures. Basically anything that trades on a futures…
6/n
contract they’ll capture momentum and other well know futures effects like spot-futures relationships etc. They make for a decent alternative asset from a capital allocation standpoint but are hardly immune to market drawdowns and 08 saw momentum on equities get hammered…
7/n
Maybe not so bad for CTA, but still REALLY bad performance. Either way that just a bit of insight into the role of these things and their flow in general. That way when you hear CTA you don’t think commodities fund you think futures momentum/ crowd effects everyone knows about.
La fin
Read this again and the grammar is so bad. Hopefully it’s all somewhat comprehensible
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Another cool flow idiosyncrasy that may interest those flow nerds out there. Cryptocurrency markets have more momentum in the short term, which is usually dominated by mean-reversion effects. This is basically all driven by over-levered degens and algorithms that...
1/idk lol
get access to these 100x leverage futures or 20x levered futures, and unlike most would think it actually gets used. Especially if you are some sort of HFT algorithm and now you can lever up as much as you want in a super volatile market. Plus with the sorts of...
2/n
profits some of these guys make, who wouldn't lever this thing up to capacity and get in while it lasts. I will probably still attribute most of the momentum still to retail guys getting liquidity cascaded out of positions. i.e. takes loss, gets out, price moves down...
3/n
Large quantity of unique features
Really good dimensionality reduction
Ensemble everywhere!
A word on each...
When it comes to modeling everyone always goes to their favorite NNs like LSTMs etc or LGBMs and those are great, but everyone has them, and frankly, they aren't that hard to implement! Just look at Kaggle if you want an example of DS students using them everywhere...
For real alpha, you need to focus on the three most ignored areas (there is a fourth, speed, but that's not really modeling, and a fifth which I'm not telling you because I like my alpha unleaked). That sounded super guru-like, but I promise these work and I use them.
For those wondering if they are, I'll give a few comments:
T-SNE is always something I want to apply, but can never quite figure out the right way to do it. There is certainly some benefits to be had from having a basic understanding of what this all means so you can get a better chance at visualizing your features in 2D...
Much like stochastic methods, as much as I would never make them the center of a model, there is always use as a feature or ensemble. Ensemble is truly the free lunch of alpha...
A key concept for MMs is how you manage inventory. Avellaneda and Stoikov is basically the model everyone uses for this. Then there comes the offset, basically how wide your spreads are. That's your basic model of liquidity provision...
From there we get to have some fun! If you can create multiple forecasts for different timeframes (and at a super-advanced level compute speeds) you can make spreads asymmetric and intentionally hold inventory...
Entirely unprompted here, but please check out @FadingRallies. Also @choffstein's Liquidity Cascades paper (link below). The flow between MMs, passive funds, ELS, and generally the effects of reflexive dealer hedging are key to understanding this regime!
Even if you aren't a trader (I certainly am not, although I try to keep up with it all) it is still super important to understand the regime and how it all fits in from a risk perspective. You CANNOT just take the models as your risk! Eigenportfolios decay, I would know I work
with them all the time so that isn't even the perfect metric (although I do love them). Statistical models will capture some risk but at the end of the day, you choose the parameters and the distribution you feed in is key. Knowing fat tails exist is incredibly important for this
Tweeting a question I was asked/ response regarding MM:
(me adding bonus resources):
A great example of C++ HFT MM algorithms. An improvement idea I have suggested to the author but can also be attempted by interested algotraders is that a fast model like XGBOOST (there is a C++
library) is used alongside some alphas to make spreads asymmetric before traders can trade against you and you get negative edge in those trades. A large part of market making is cheaply executing alphas by trying to get inventory on the side of your predictions and also by
getting out the way of adverse conditions by making your spreads asymmetrically wide (traders with alpha against you). github.com/hello2all/gamm…