“When we try to pick out anything by itself, we find it hitched to everything else in the Universe.” – John Muir, Mountaineer
It’s one thing to build models and share insights about specific players in the trading and asset management universe. It’s an entirely different thing to perform comparative analysis of that specific modeling to develop various rankings of a community of players. This latter point is precisely where the accumulation of Alphacution’s modeling and research is now taking us, and that new level of insight is, frankly, a bit mindblowing.
Our previous Feed post that illuminated an apparent anomaly with Jane Street’s stock selection strategy was one of our first examples. For this one, we look at a ranking of average stock positions by shares for a selection of leading quant hedge fund managers, market makers and proprietary trading firms. Now, what’s truly fascinating here is if you consider these players in order of average stock position (by shares) you realize that what you are simultaneously looking at is a ranking by strategy turnover frequency – or, in short, strategy speed.
Let’s find out…
Here’s the same exhibit presented in log scale to spread out the data:
The basic logic behind our claim that average stock position in shares is indicative of strategy turnover frequency is this:
Market liquidity is the key limiting factor – along with explicit trading costs – for how much “edge” can be harvested without degrading performance. In other words, market liquidity helps define strategy capacity. Furthermore, faster strategies – those with high turnover and short holding periods – tend to have lower capacities. Think of it like this: The more a strategy needs to interact with market liquidity with specific timing, the lower its capacity is likely to be. Also, faster strategies tend not to accumulate much in the way of positions because their outsized performance is dependent precisely on the fact that they hold the minimum amount of risk.
Therefore, one way to expand the capacity of fast strategies is not to build positions but to expand the roster of products that are subject to the strategy. Alternatively, slower strategies tend to have higher capacities (for the opposite reason as stated above) and tend to accumulate bigger positions. Here, too, the capacity of slower strategies can be further expanded by expanding the roster of products targeted by the strategy.
So, generally speaking, the more/less capital you need to put to work in the market (while maintaining performance expectations), the slower/faster you will likely need to trade – and the bigger/smaller your average position is likely to be…
Now, go back to the charts and see if this logic makes sense to you.
Either way, send feedback…
One last thing: This analysis is like opening Pandora’s box. There will be a ton of room for debate around the edges. For instances, most of these firms are multi-strats. How do we account for the impact of that on average turnover frequency? And, any firm can purposefully trade small. Have we taken this into consideration? Certainly, there are numerous other factors to deconstruct and defend. We have arguments for many of these pushbacks that we can lay out in a longer post or deep-dive case study at some point…
This is simply a first pass at the idea that there is an empirical method available to measure and rank some of the most mysterious and mythological phenomena in the global markets ecosystem (such as how Renaissance figured out how to trade so slow on an automated basis). In time, and with further development of the modeling – including certain position level liquidity modeling – we may be able to assign actual average durations to each strategy.
Who knows? We never know what we will learn as we embark upon on each new puzzle-solving adventure. We only expect that we will learn something new and useful along the way which can be converted into intelligence for market actors and their stakeholders…
Support the Feed!
Note: Business credit cards and bank accounts can be used via our PayPal payment portal.
Alphacution is in the intelligence business.
For those of you who are eager to derive greater value from this work and apply that intelligence to your own business interests, Alphacution is offering unaffiliated individual subscription options priced at $275 per year or $25 per month, cancellable at any time. Both of these options include a rebate on purchases of deeper, more substantive reports and case studies.
In other words, the entire value of an individual subscription paid up to the point of purchasing a single report will be deducted from the purchase of that report. (Rebates not to exceed the maximum value of an annual subscription.)
Enterprise subscription packages for individuals affiliated with trading firms and custom content/service engagement options are available upon request at email@example.com.
Now, for those of you who don’t expect to take advantage of the offers outlined above but want to continue to enjoy the insights, intelligence and occassional entertainment that remain openly available on the Feed, I want to make this specific plea:
Free doesn’t mean there are no costs. In fact, in this case, there have been extraordinary costs in the accumulation of experience and sight, meticulous curation and assembly of data, and creative visualization of and storytelling around our findings.
So, if you value quality content – here or anywhere else – then you need to find a way to support that content at some level simply because you want it to continue to exist. Our post, In Support of Digital Content – which was adapted from other notable digital era content developers – makes a more expansive case for this perspective.
Bottom line: Your efforts to support via one-time or recurring contributions will help guard against this content needing to move from the currently preferred audience-driven model (for its level of independence) to a sponsorship-driven model (which can be found on most other industry media outlets).
So, if none of the subscription options suit you, one-time and recurring support contributions can be made at any level here:
Of course, as always: If you value this work, please continue to “like it,” share it, comment on it – or discuss amongst your colleagues – and then send us firstname.lastname@example.org.
As our “feedback loop” becomes more vibrant – given input from clients and other members of our network, especially around new questions to be answered – the value of this work will accelerate.
Don’t be shy…
Unsubscribe from prior subscriptions without further obligation, at any time, here: