“Do not let what you cannot do interfere with what you can do.”
John Wooden
In the Introduction section – downloadable for free by all registered subscribers – of Alphacution’s recently published case study on US Equity Option Market Makers, we present an outline of the two-factor modeling method used to estimate option volume share by market maker. The chart below shows the key product class benchmarks used as guideposts for those estimates where open interest is used in the denominator of the trading velocity ratio…
One of the most fascinating discoveries from this study – and one that unlocked numerous other discoveries and hypotheses contained therein – relates to Chicago-based option market maker and asset management group, Wolverine Trading. Wolverine is one of five core wholesale option market makers for retail option order flows; the others being Citadel Securities, affiliates of Susquehanna International Group (SIG), affiliates and agents of IMC Financial Markets (IMC), and Morgan Stanley & Co. (MSCO)…
Alphacution estimates market maker share of retail order flow volume in stocks and options via regulatory data. Based on that data, Alphacution is able to further estimate the monthly trading velocities for each of these five wholesale option market makers. Given our growing sense of the core option market making strategies for most of these players, plotting their weighted average annual trading velocities made sense. However, when we plotted trading velocity for Wolverine, it did not make sense – at first. The chart below highlights Alphacution’s estimate for Wolverine’s trading velocity based on its wholesale share of retail order option flows relative to those product class benchmarks…
Here, this chart is showing that Wolverine’s retail option order flow is turning over faster than SPY options for 2019 – 2021, and roughly 2.5 times faster than the stock option benchmark for 2022 and 2023. Now, with our 2019 estimates being less credible than the other years, any conclusions for 2019 need to be more heavily discounted. But for the other years, 2020 – 2023, our estimates are fairly solid; leaving our estimates for monthly long open interest (from quarterly data) along with total estimates for short open interest as the main weak points in this methodology. However, given that option market maker books are rarely tilted more than about +/-10% (net fair value as a percentage of gross fair value), Alphacution doesn’t expect that even full transparency on open interest would impact our trading velocity estimates for Wolverine all that much…
The chart below cleans some of this up by illustrating both monthly and annual trading velocity estimates for Wolverine – shorthanded to WEX (which is actually the acronym for affiliate Wolverine Execution Services) – and stock options as a group…
The chart below is the same as the chart above except with the stock option trading velocity placed on the the right hand scale (rhs) to provide an alternate visual comparison for when Wolverine is trading with the stock option group and when it’s trading differently than the stock option group…
Anyway, Wolverine’s trading velocity estimate didn’t make sense at first largely because its 13F reports don’t support large positions in SPY options – and certainly not relative to the SPY option positions at CitSec, SIG, and IMC. At first, we concluded that activity in SPY options would be among the most likely explanations for trading velocities displayed by Wolverine’s retail curve. So, things didn’t add up…
As we further developed the case study, it clicked – as it may have already clicked for you: Wolverine’s peak trading velocity in 2021 is the key clue because it roughly corresponds with the meme stock craze. And based on this hypothesis, Alphacution was able to develop a broader hypothesis – with diagrams – on the retail option allocation scheme to each wholesale option market maker – with Wolverine being allocated orders in traditionally lower liquidity, sometimes out-of-favor names like AMC Entertainment (AMC) or Bed Bath & Beyond (BBBY); the exact universe of names that meme stocks come from. Note, of course, that our case study goes into much more detail about this…
Now, for this Feed post, we decided to go beyond the case study and test an additional hypothesis, presented for the first time here: Wolverine is one of several option market makers with lower-frequency, asset management affiliates. In this case, that affiliate is Wolverine Asset Management (WAM). So, we thought perhaps the high trading velocity at Wolverine might have something to do with option order flow that was passed through to WAM, thereby skewing the trading velocity ratio for the combined group. In other words, we needed to count an estimate for WAM’s option open interest in the overall Wolverine trading velocity. The chart below adds both monthly and annual trading velocity estimates for “WEX + WAM” to the prior chart format…
Finally, the chart below adds WAM’s estimated open interest percentage to the right hand scale plus Alphacution’s estimate for CitSec’s estimated option trading velocity to the prior chart format…
Based on the depth of Alphacution’s analysis so far, WAM’s option open interest is a small part of the total (~6% – 18% during the 48-month period), and therefore, doesn’t lower Wolverine’s overall trading velocity much. In other words, whatever pass through may be occurring at Wolverine, it’s not having much impact on trading velocity…
So, we’re going to stick with our original hypothesis that Wolverine is receiving retail option order flows in the subset of the optionable universe that includes meme stocks until or unless we develop a better hypothesis…
Until next time…
