“Two wrongs may not make a right, but a thousand wrongs make a writer.” – Dennis Miller
“If you’re not making mistakes, then you’re not doing anything.” – John Wooden
In my last in-person meeting of the B.C. (Before Coronavirus) era, a noted Chicago hedge-fund-slash-market-making C-suiter told me: “You’re just making this stuff up…”
Answer: “Well, yes – yes I am. Based on an interpretation of an assembly of data supplied by you – and others…”
There is no roadmap here, other than the one that is constantly being course-corrected in my head. No textbook. No lexicon. No Ivy League curricula to provide structure or Council of Elders to guide. Just the kind of seemingly unending puzzle-solving exercise to soothe a delusional level of puzzle-solving desire… And, the feedback loop supplied by you – Feed readers, subscribers and clients, alike – that informs a faint sense of navigation.
With that, “we” are just making stuff up: Inventing language and pictures – and the stories that seem to fit well with them…
And so, it is with this brief yet indulgent setup as backdrop that we arrive at the heart of the matter: Sometimes, the stuff I make up – uh – is wrong.
Now, I can play a bit of jujitsu – and deflect the cause of mistakes to the wee-ness of the hour, the messiness of the data, or complexity of the model. But, I think the bigger value here is in confessing the creative and uncertain nature of the Alphacution road. Thereby – ideally – forging a partnership with you; that we are swinging from vine to vine in a jungle of data that very few have ever tried to tame – and, by the looks of it, no one has been foolhardy enough to try to tame out in the open…
Inadvertently, mistakes will be made. On best efforts, I will correct some of those here, like this: In a Feed post from July 2019 entitled, “Jane Street’s Twist on Stock Selection,” Alphacution performs a comparative analysis of average implied stock prices for a selection of legendary trading firms based on a ratio of aggregate 13F value of product class exposures and total 13F product class positions, in this case 13F (long) stock holdings. In that analysis, Jane Street was notable for its divergence from the sample, below:
Now, chances are, the numerator data was grabbed from the wrong field to produce such an aberrant result. And, yes, normally I check and double-check outliers. And, yes, maybe in this case, it was the wee-ness of the hour or the “mile of the marathon” – but, that doesn’t change the fact that it was wrong…
As we gear up to publish Alphacution’s next comprehensive case study on Jane Street Group, the relevant models – including those of our prior case study targets (Citadel, Susquehanna and Two Sigma) – have been upgraded and updated. Turns out, though Jane Street certainly has its own twist on stock selection, it still tends to run with the pack. See below…
The good news – and part of the aforementioned roadmap – is that there are more signals to be harvested from this data even though it’s flowing at a slow update frequency. With increasing automation, we should be able to figure out more detailed snapshots of security selection criteria – and strategy variances among key players. Furthermore, we should be able to probabilistically construct (estimates of) the short sides of these books, as well.
I haven’t figured out precisely how to swing on that vine yet. It’ll likely happen when “we” is more We than I… And, once we get there, who knows what else we might see?
Until next time, please stay safe out there…
Support the Feed!
Alphacution is gearing up to put more value into its basic individual subscription, including exclusive content. To support the Feed in anticipation of these upgrades, click Subscribe below. For details on basic and premium subscription offerings, send inquiries to email@example.com.
Note: Business credit cards and bank accounts can be used via our PayPal payment portal.