Authors Macrocephalopod
7 days
30 days
All time
Recent
Popular
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.
This is a problem because the costs of running a pure MM firm (mainly infrastructure and employee comp) are increasing and profit margins are decreasing. So many firms turn to prop trading as a way to increase P&L at the cost of some Sharpe.
One way to approach this is to make your price skew dependent on factors other than your inventory, eg if you think the market is going up you skew prices a little higher to encourage people to sell to you and discourage them from buying from you.
This is important to realize. The way many of these firms are structured is that they have a pure MM book (keep inventory low, stay hedged, clip spreads) and prop books to express a view. The prop book can trade with the MM book at (to first approximation) mid and a lot of their
— macrocephalopod (@macrocephalopod) February 15, 2021
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.
This is a problem because the costs of running a pure MM firm (mainly infrastructure and employee comp) are increasing and profit margins are decreasing. So many firms turn to prop trading as a way to increase P&L at the cost of some Sharpe.
One way to approach this is to make your price skew dependent on factors other than your inventory, eg if you think the market is going up you skew prices a little higher to encourage people to sell to you and discourage them from buying from you.
A few things that I didn't cover yesterday when I talked about equity factor models (it's a huge area and it's impossible to more than scrape the surface)
1. How do you get the exposure matrix Xt?
There are different ways to estimate it, depending on the factor. Simplest is factors like industry or country exposure where the entries can be 0/1 depending on whether the stock is in that industry/country or not.
Some exposures can be estimated by linear regression on historical data, if you already have a time series which approximates the factor returns. E.g. exposure to the market factor (beta) is estimated this way, by regressing each stock against the S&P 500 (or some other index)
This also works for "macro" factors e.g. you can estimate exposures for each stock to commodity prices, exchange rates, interest rates, GDP or inflation surprises etc by regressing stock returns against the relevant historical time series.
Finally you can have exposures which are heuristically derived from other observable data about the stock, e.g. accounting data, analyst reports, past price movements etc. In this case you find some metric which measures the factor you care about (e.g. price to earnings) and
A few people in the DMs asking about equity factor models so here's a short explainer.
— macrocephalopod (@macrocephalopod) February 2, 2021
Let's make it a concrete problem -- you are the risk manager at a big multi-manager hedge fund with ~100 sub-PMs each of whom has a portfolio of 10-50 stocks, long and short.
1. How do you get the exposure matrix Xt?
There are different ways to estimate it, depending on the factor. Simplest is factors like industry or country exposure where the entries can be 0/1 depending on whether the stock is in that industry/country or not.
Some exposures can be estimated by linear regression on historical data, if you already have a time series which approximates the factor returns. E.g. exposure to the market factor (beta) is estimated this way, by regressing each stock against the S&P 500 (or some other index)
This also works for "macro" factors e.g. you can estimate exposures for each stock to commodity prices, exchange rates, interest rates, GDP or inflation surprises etc by regressing stock returns against the relevant historical time series.
Finally you can have exposures which are heuristically derived from other observable data about the stock, e.g. accounting data, analyst reports, past price movements etc. In this case you find some metric which measures the factor you care about (e.g. price to earnings) and