The DARWIN Exchange (Darwinex) lists DARWINs that investors purchase from independent traders. To assist investors, we grade DARWINs for investable attributes (IAs).
But: DARWINs are (different leverage) transformations of their underlying strategies. So… what are the IA grades for, the DARWIN, or its underlying strategy?
The logical answer would be – the DARWIN! Then again, remember Tradeslide and the TS Score? Way back when, we were a site that diagnosed trading strategies, and DARWINs didn’t exist. Because of that legacy, up to last week, IA grades – somewhat inconsistently – were for the underlying strategy, not its DARWIN.
For the Consistency & Timing IA grades this is irrelevant: a strategy’s and its DARWIN TP / SL levels and trade timing are identical. For other IAs DARWIN & underlying strategy differ.
To side with logic, going forward IAs will grade the DARWIN – not the underlying strategy. The rest of this post outlines the implications for the Experience and Performance IA.
1. Changes to Experience Attribute
As you may recall, the Experience IA measures statistical representativeness. One core input into it is, amongst others, D-Leverage.
To rate the DARWIN’s statistical representativeness (instead of the underlying strategy’s), the Experience IA now feeds off DARWIN adjusted D-Leverage, as opposed to underlying strategy’s D-Leverage. After this change, which track record is more statistically representative, a strategy’s, or its DARWIN’s?
To reach the answer, think this through. All other things being equal, which strategy has more statistical significance?:
- Strategy 1: Take 100 trades, one where you leverage 200:1 for 5 days, and 99 where you leverage 1:1 for 1 hour
- DARWIN 1: Strategy’s 1 DARWIN, as offered to investors
Strategy 1 is (probably) very driven by the outcome of the 5-day 200:1 trade. To protect investors in DARWIN 1, risk management algos would not fully replicate the 200:1 leverage trade to begin with. Moreover, as the 5 days progressed, they would gradually close out the trade to control leverage over time. This “smoothing out” would increase DARWIN 1’s statistical significance above Strategy 1, by “allocating” performance (more) evenly across the remaining 99 trades.
DARWINs accrue Experience at least as fast as their underlying Strategy, and generally faster. Going forward, therefore, users should rejoice that DARWIN Experience grades will grow faster – precisely because it rates DARWINs, not underlying strategies. Note that investability doesn’t change one bit – it’s just that the grade refers now to something else.
2. Changes to the Performance Attribute
Similarly, the Performance Investable Attribute grades the output DARWIN’s performance (i.e. the performance on 20% VaR after algos manage D-Leverage independently from the underlying strategy).
As you may recall, the Performance IA used to include a Leverage Illusion grade in addition to the Monkey Test. Their average grade was the strategy’s 0-10 Performance score.
Because only adjusted D-Leverage drives DARWIN performance , the “with and without leverage” Leverage Illusion grade is now redundant.
All DARWIN providers need to do now is beat trading monkeys that use their same D-leverage!
Think this makes beating the monkeys any easier? Think again 🙂
3. Changes to other Attributes
Evolution is a process, so it doesn’t stop here! We’ll keep working hard on algorithms and ratings.
Questions on the changes or anything else Darwinex?
E-mail email@example.com, we’ll be glad to help out!