Hello Forum. I would like to ask for your professional opinion about if there are good numbers for the return and Sharpe, which make a strategy "interesting". Background: we are a small team of Ph.D's and engineers, who developed a trading platform, which Is based on the "reading and understanding" of the market microstructure dynamics. The previous version of our platform was developed to maximize the profit, but even though we achieved an average 82% probability of profitable trades, a small subset of trades was producing large drawdowns. As we are currently finalizing the next version of the platform, in which we completely rebuilt the risk control module, I would be interested in your professional opinion is there a "sweet spot" between a "Low profit/high Sharpe" and a "High Profit/low Sharpe" strategies, and if there is one, what are the numbers we need to achieve, or to beat? I know that important also is the capacity of the strategy, but as we work with the US index futures (mostly ES), the scalability is not an issue for us. Thank you.
Trending Content - Hedge Fund Forum
| +38 | How much modeling do experienced HF analysts/PMs actually do? | 13 | 7h |
| +21 | Buy Side Credit - Seats / Guide | 8 | 11h |
| +18 | London buy side | 2 | 6h |
| +13 | Hedge Fund Pivot | 4 | 6m |
| +11 | Mid-career shift to buy side with niche background | 3 | 3d |
| +9 | Wfh / work remotely at MMs or SMs? | 5 | 3d |
| +3 | Atlantic Wolf Capital Closing? | 5 | 3h |
| +3 | How old is too old? | 3 | 1d |
| 0 | Hedge funds | 1 | 3d |
Career Resources
Based on the most helpful WSO content, the preferred ranges for Return and Sharpe in portfolio management depend on the type of strategy and the expectations of the investors or institutions involved. Here are some insights:
Sharpe Ratio:
Return Expectations:
Trade-offs:
Drawdowns:
Sweet Spot:
Your focus on US index futures (ES) and the improved risk control module is promising. Ensuring that your strategy maintains a high Sharpe ratio while controlling drawdowns will make it more attractive to institutional investors.
Sources: Is the Multi Manager HF Experience Worth It?, Do I have a good strategy?, MultiManager PM (Millenium, Balyasny, Schonfeld) minimum requirements, Diminutive Nature of Net Returns, Hedge Fund is Paradise
i am still a student, so correct me if im wrong.
shouldnt portfolio managers look at alpha and information ratio rather than sharpe ratio?
since they measure against some benchmark; if that benchmark gives abnormally large return in one year, that would increase its std deviation hence lowering sharpe ratio. then the pm would get punished for following the benchmark by lower sharpe ratio.
While you're not wrong, in reality, the industry standard is to still to look at Sharpe. If you have a strategy which has a low Sharpe due to large upside deviations, then it's possible to explain this using the Sortino Ratio..
Maybe at non-L/S funds. L/S, and particularly at pod shops, is exactly designed to run market neutral and not get benchmarket
Based on the information you provided, I gather the strategy you've built is a short-term mean reversion style strategy. The profile of these types of strategies exhibit high win-rate, low profit/loss ratio and thus have negative skew (short volatility). While this isn't necessarily an issue, it just means you can achieve periods of high Sharpe, but then have infrequent, disastrous events that can wipe out a significant portion of your capital. The difficulty with this is managing/knowing the extent of these negative events.
Generally, a Sharpe >2 is what you should be aiming for, but this is without knowing anything about the strategy. The profile of strategy impacts what Sharpe you should be aiming for though. Assuming you have a negative skew strategy (short volatility), then I would suggest a higher Sharpe, maybe 2.5 - 3, this is so you can be comfortable with the recovery/buffer from adverse events. If you have a positive skew strategy, then you could get away with a Sharpe in the region of 1.5 - 2, this is because your strategy is likely long volatility and can take advantage of significant market events/moves..
Thank you, PM in HF. You are quite right suspecting a flavor of a "short-term mean reversion". This is exactly how the platform started. I have to admit that the early success let me think that this is how it would continue its operations, but with adding more instruments, especially the ES-related "family", especially in the world, of the 0DTE-based strategies, and all other complications (like HFT, e.g.), led to those "disastrous events" that you correctly identified. Thus, the efforts that I mentioned as "completely rebuilt the risk control module", included a complete redesign of the pattern sequence analysis methods (we collect tick data and transform them into much lengthier patterns, that reflect the liquidity dynamics as well as the buy/sell activity). I probably do not completely understand if the "negative/positive skew", is quite applicable to the strategy, as the idea behind it is to be "agnostic" and let the liquidity dynamics, caused by "large and sophisticated" liquidity providers/consumers, "dictate" the next play. In other words, we can go long or short, depending of what side (buying or selling) feels "weaker". I would very much appreciate it, if you please could elaborate a bit more on the differences between those two strategy types, as I really would like to understand what I, as a physicist, do not know that I do not know :) Thank you.
Yeah, no problem. The "skewness" of a strategy just refers to a moment of the PnL distribution. If your strategy has negative skew, it just means you have a left-skewed distribution with a tail. You have asymmetry in your distribution due to some outsized losses. Positive skew is the opposite. I'm sure you would've covered these basic statistics based on your physics background. Now, the relevance to your strategy, or strategies in general, is that every strategy has a skewness profile. You can broadly interpret skew in two ways, short volatility or long volatility. If you think of a generic short volatility strategy e.g. selling a straddle; most of the time (high probability) you will make small [fixed] sized wins but every so often you will have a volatility event which produces large [unfixed] sized losses. This type of options strategy generates a negative skew profile. By knowing this, you can then class your strategy as either "short volatility" or "long volatility", indicating how your strategy performs during significant market events / volatility events.
The point of all of this is to identify your strategys profile and to manage the risk accordingly. Both categories of skew have their pros/cons, and you need to understand them to best position your strategy.
Interpreting the above comment, basically if you’re “long/short” ‘depending on what is weaker’ is wrong, how fast are you able to tell this, decide to get out, and what is the implementation shortfall to exit (your left tail), and how often will this occur vs. being “wrong” to the upside (if it’s mean reversion then say you’re wrong on the speed, so can exit faster, or it overshoots and you can reliably capture the overshot move) which is your right tail.
Then the ratio of the two tails across different regimes should tell you what risks you’re are actually being compensated for.
you’ve probably heard of the lore around hubris, quant strategies and leverage embodied by Long Term Capital Management which would run mean reversion and harvest the Illiquidity premia on treasury bonds/futures.
My own advice (take or leave it), is if you’re able to have very good analytics/predictive abilities for your largest drawdowns (capturing regime changes), that’s the best defence against hubris. What I mean by that is if you can model how wrong you are to 99.99999% probabilities (not hyperbole - 5 9’s is probably as robust as you will need long term, especially if using leverage) and the damage it does to your returns, then you can understand how much return you need to generate (and whether you are in fact creating alpha (which is zero sum; i.e. you are simply a better market disseminator of information and are rewarded as such basically like a commission, hidden liability free), or are simply collecting a premium for tail-risk insurance in the markets you operate from these other market players (I.e. this is not alpha, but really a risk premia that market participants choose not to own and are paying you to own) - which is still fine, in that sense you are acting as a specialist insurer, therefore you should have sufficient reserves across time for the hidden liability you are accruing for the inevitable payout during a regime change).
put another way, know which business you are in because it dictates whether you have a hidden liability or not,
Et omnis rerum soluta qui quia laborum. Maxime iusto aliquid magnam est fugiat voluptatem voluptates.
Delectus occaecati in id voluptas et repellat. Aut nemo tenetur molestias magnam. Tempore omnis unde aliquid autem error. Non officia sunt fuga. Officia illum dignissimos aut. Et ut assumenda et blanditiis et.
Quaerat reiciendis dolor dolorem et esse minus. Ex dolorem amet unde nemo est rem officiis quo. Non earum velit est dolores et adipisci sit. Eum quod eius autem. Quam commodi fuga eum qui consequatur quae. Id inventore hic perferendis totam.
Tempore architecto temporibus eaque ducimus saepe. Ad repellat nam laboriosam voluptatem delectus perferendis sed. Voluptatem eveniet qui sunt unde rerum.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...