People love positioning/flow arbs on this site (dealer option hedging, target vol and risk parity flows, convexity hedging etc etc, all the same shit with different names) so let me tell you about the **greatest flow arb that ever existed**
The strategy had double digit return and high Sharpe for about eight years from 2000-2008, with a kicker that let you continue running it until ~2014 if you were crafty
First you need to know about the GSCI -- the Goldman Sachs Commodity Index. It was created by GS in 1997 to track passive long exposure to commodity markets. Crucially it is an *investable index* which means that you can replicate its returns by trading in liquid instruments,
in this case commodity futures. GS fully laid out the methodology for calculating the index and how to replicate it by trading in futures, which meant that dealers were able to sell swaps on the index and pull in a nice fee, which they started doing in size.
The index is now administered by S&P and you can find the full methodology here. Apart from some tweaks to the exact commodities included and their weights, it is exactly the same as it was in 1997 spglobal.com/spdji/en/docum…
The key that makes the trade possible is to replicate the index dealers need to hold near-term futures and roll them to a later expiry before they become due. The index methodology lays out the exact schedule of rolls for every commodity, including which contract months will
be held, and which days they will be rolled on -- in basically every case the roll takes place over five days, with 20% of each position being rolled before the close on the 5th, 6th, 7th, 8th and 9th and ninth business days of the month.
Because the aum tracking the index got so large, this created a predictable and sizable flow in commodity futures spread contracts which you could get ahead of. The strategy was to put on the same spread that the dealers would need to do, five days before they
needed to do it, and then sell the spread position to them as they begin the roll. Doing this equal-weighted across all futures in the index gave you a strategy with ~25% returns and a Sharpe of 2.5 from 2000-2008 --
After 2008 too many people knew about it and the trade became unprofitable after costs (and newer commodity indices were developed which weren't as vulnerable to front running) but a cute twist was to try to front-run the front-runners, i.e. do everything another 5 days early
This wasn't quite as profitable but it meant you could continue doing the trade for another six years or so, until it finally died out around the end of 2013.
Trying to repeat the same trick didn't work -- by trying to push it earlier and earlier, you end up "running into" the previous month's roll and it's no longer profitable (plus by this point, anyone who was anyone knew the idea and was probably trying it)
No big lessons from this but a cute example of how fixed, systematic trading strategies can create price distortions when they get large. I haven't looked but I'd be willing to get you can find similar effects created by some of the big "smart beta" funds today. FIN.

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with macrocephalopod

macrocephalopod Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @macrocephalopod

19 Feb
I wrote here about a misconception people have about float-weighted funds, namely that they "distort the market" by mechanically buying the stocks that rise the most, which is NOT TRUE, but I think there *is* something interesting to say
Approx one million people wrote me to say that it's not about the funds re-weighting, it's about the FLOWS i.e. when the price of a stock goes up, the fund buys a larger $ amount of it the next time it receives an inflow. This sounds convincing but is also wrong ...
for the simple reason that if a float-weighted fund needs to buy e.g. 0.1% of the market due to a new flow, it simply buys 0.1% of the shares of every company. If you want to claim that this distorts prices you need a theory for why buying 0.1% of a companies shares causes more
Read 15 tweets
18 Feb
This is true, breakeven inflation = expected inflation + risk premium - TIPS liquidity premium, sadly you do not get to observe the last two so... 🤷
Basically every observable metric derived from asset prices has this problem, e.g. treasury yield = risk-free rate + term premium + expected change in yields - convexity premium, only the first (and maaaaybe last) are observable.
Equity dividend yield = risk-free rate + equity risk premium - expected dividend growth rate, the last two are not observable.
Read 5 tweets
17 Feb
Say you want to execute a $30k trade in a $1bn mkt cap stock, around 0.2% of adv so not massive but not small. Would that be cheaper in the US (recognized as the most liquid stock mkt in the world) or in South Korea (which has a 10bp transaction tax)? linkedin.com/pulse/which-co…
Surprisingly it doesn't make much difference -- you would pay ~25bp (or $75) in both markets, even with the transaction tax in SK. In Japan you would only pay 10bp ($30) to do the trade!
The difference is that US equity markets are fragmented (~13 different exchanges) whereas in South Korea everything trades on a single exchange, KRX
Read 8 tweets
15 Feb
Some thoughts on how big market making firms (eg Jane Street, Susquehanna, Optiver) are structured. Note I have not worked at any of these firms so this is not based on any insider knowledge, just talking to people in the industry and extrapolating a bit.
A “pure” market making operation is based on clipping spreads, ie buy low, sell high, keep inventory low, keep risks (eg greeks) tightly hedged. Skew your bid/offer based on your inventory to try and offload it as quickly as possible without impacting your profit too much.
This kind of trading has enormous risk-adjusted returns (Sharpe > 10, ~no down days) but it’s hard to scale it because your P&L is a function of two things — volume and volatility — that you don’t have any control over.
Read 12 tweets
11 Feb
A wrong belief that many otherwise smart people have is that index funds are “momentum investors” i.e. they mechanically buy more as prices rise and sell when prices fall (example from Bill Ackman’s 2015 investor letter below and I saw this *twice* in unrelated threads today). Image
This is not true! A really basic cap-weighted index fund mechanic is that they automatically track the index as prices move, with no trading required. This is literally the reason that they are called *passive* funds.
For some reason this breaks people’s intuitions and they get really excited about forced buying/selling from passive funds creating positive/negative feedback loops, which, I cannot emphasise enough, is not a thing that actually happens except in retail brain fantasies.
Read 9 tweets
9 Feb
Options greeks ranked from worst to best, by someone who doesn't really understand options.

10. Rho -- worst sensitivity, very boring, mostly ignored except by people trading very long maturities (who you should not trust anyway)
9. Delta -- very terrible and boring first-order sensitivity. Hedged by everyone except retail. May be driving the stock market now????? (unclear)
8. Theta -- seems sexy and interesting at first, really just the opposite to gamma upon closer inspection. Has potential to blow up sellers (only reason it is ranked above delta)
Read 10 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!