Being curious at the returns achieved by bank prop trading desks, I did a search on Companies House beta (UK govt website) on the prop trading subsidiaries of BNP Paribas & Societe Generale, which are foreign banks that continue to have prop trading in the post-Volcker rule era. A few details as follows, BNP Paribas - Prop trading arm: Opera Trading Capital Total of 21 traders (according to FT article last year) As of Dec 31, 2017, accounts on Companies House show trading profits of 32.9 million euros, on total shareholders' equity (ie. capital) of 600 million euros. The figures for 2016 and 2015 are 114.9 and 32.4 million euros, on roughly equivalent capital bases. This works out to about 9.5% annual return average for the 3 full years since the records started. Barclays spinoff - Squarepoint Capital Avg employees in 2017: 65 Trading profit: 46m pounds in 2017, 26m pounds in 2016 AUM of $350m reportedly Societe Generale: Prop trading - Descartes Trading (fully owned by SocGen) Trading profit: 47.8m euros in 2017, 23.7m euros in 2016, -9.4m euros in 2015 AUM of 371m euros, according to IFR So it would seem that the returns are <10% annually on average for the prop desks, which makes them on par with the avg mutual fund/ buy and hold long term investors. These figures are somewhat lower than i'd been expecting, with all the secrecy and presumed "savviness" that surrounds bank prop traders. Are these returns % indicative of bank prop desks in general? (eg. before Volcker rule in US, are the top US banks making roughly equivalent % returns) or maybe it's just the French traders who are posting low teens returns On a sidenote, although much has been made of the low 'success' rate of retail traders, i think it's probably got to do with the % return targets - ie. if retail traders merely targeted 10% per year like the bank prop traders, they can probably achieve it too, but that option seems unappealing to many because of the low absolute $ that entails. & thus the anecdotal 'success' rate of bank prop traders seems high - ie. you hardly hear of guys really being kicked out bc they blew up, mostly it's just reshuffling across different trading desks because of higher pay packages, etc, bc their annual % return range is most likely in the -5% to +10% region. For extra credit, what educational materials do banks provide for their traders? eg. I read the Lehman FX training manual, but it was nothing too exciting.
Mutual funds are not that high. 1997 to 2017 average annual mutual fund returns are 4.67%. ~10% return is great, if it was consistently achieved, while trading different less correlated products to the S&P500. Too little years (2-3) in the data you found though to make a conclusion.
i guess it depends on how you select the time window for Vanguard S&P 500 based mutual fund, past 10 years return at around 10.7% per year: i mean, you could dice it by beta, mkt exposure, etc, to justify fee structure paid by clients of a hedge fund, but if we're talking about a business from the vantage point of shareholders' equity, ultimately it should be ROIC
I spent a good chunk of my institutional career on a prop desk for a large bank. These numbers seem very low. With 21 traders I would have expected the gross profit to be around 200mm+. these statements feel like the balance sheet is managed to produce zero pnl. And the 21 traders are trading through various entities. We certainly did and as traders we didn’t care where the pnl resided but some tax attorney at the bank probably did. We never saw our balance sheet costs but one of the other internal prop groups spun out and the the audit on their track record indicated a 19percent annual return over 5 years.
thanks for the input. just wondering, to generate profit of 10mm per seat, what would be the trading limits/ capital allocation for the avg trader?
I have forgotten what the limits were. There was no concept of capital allocation. You were charged for your balance sheet usage and only constrained by risk limits such as delta, Vega, and various gap risk measures. This was for an equity volatility book. I don’t know fixed income desks handled tings but I got the sense they generally ran much larger books and got smaller percentage payouts.
The concept of "capital" is very nebulous because usually there is plenty of capital and your main limitation is the risk limits. What ends up happening is the firm you work for imputes the capital from your expected risk (for example, capital = 100 x var or capital = GMV/5 etc). Once you are there, they impose various forms of limits on your performance like (e.g. 5%+5% draw downs on the imputed capital or 1y Sharpe > 2) or risk metrics (e.g. max delta/gamma/vega). So it's going to differ a lot by the firm and the asset class. You can, however, roughly back out the risk limits based on desired dollar volatility and max drawdown under some reasonable assumptions.
not sure if i follow...could you make it simpler, like if a trader who specializes in long bonds comes onto a global macro desk, how many lots of ZB is he allowed to put on if trading directionally, if aiming for $10mm profit annually? like the andurand guy in crude oil, my understanding is he mostly just takes a directional view on crude oil futures, for his smaller counterpart who's part of a multi-strategy fund, what would the typical trader contract terms be like?
as an ex commods guy on a prop desk.. our limits were drawdown (~5-10mm); max var (~2mm) and profit expectation was around 1.5x drawdown on average, or 4-5x max var. There were other limits like concentration etc. All these limits were soft, if you were making a ton of money it got more relaxed and vice versa