Search results

It definitely has a seasonality pattern. Same goes for other commodities like grains. The problem is of course how to do carry trading, as per the book it is defined as a simple differential of two adjacent contracts, however this doesn't feel right for such commodities.

How do you go about measuring carry for nat gas to generate the signal? As for crude oil, do you simply look at the difference between CLZ21 and CLZ22?

Hello Robert, I've ran into some issues with regards to calculating carry. I'll try to show them with a couple of examples: WTI Crude Oil (CL) and Natural Gas (NG). Both of them have 12 active months in a year, so a monthly roll over (I know you use different scheme for your trading, like only...

Hello Robert, Your forum username is "globalarbtrader". What does it mean? Do you do any sort of arbitrage trading? (statistical arbitrage perhaps?)

That's a double coincidence that you just gave me the link to the article that I was looking at today, because I remember reading it in the past and there were some hints: My favourite setup forecast_cost_estimates: use_pooled_costs: False use_pooled_turnover: True ## even when pooling...

Hello Robert, How would you go about adjusting returns for the costs? Suppose we have cost estimates in Sharpe ratio units "c" and returns series as "series". I would do (for weekly return data): adjustment = stdev(series) * c / sqrt(52) series = series - adjustment Then these series could be...

I've done the same with automatic clustering + improved interpolation technique: The outcome was Sharpe ratio of 0.703. Basically the same as 0.711 that I got from the new recently published methodology using the SLSQP optimizer and estimate uncertainty + hacks. I guess the conclusion is: it...

If all instruments are equal weighted, the following Sharpe ratios are achieved in the backtest: automated clustering 0.711 manual clustering 0.683 equal weights 0.667

Auto VS Manual clustering weightings turned out quite different. However when checking basic stats, I'm getting: auto clustering: mean = 5.6671 stdev = 168.80 mean / stdev = 0.033573 manual clustering: mean = 5.6206 stdev = 172.09 mean / stdev = 0.032661 equal weights: mean = 4.3331 stdev =...

Do you mean applying PCA to the correlation matrix? I did not do it. The clustering is hierarchical complete-linkage algorithm with max cluster size of 3 (to accommodate the handcrafting "candidates").

Just did the first comparison: calculating weights (with diversification multipliers applied) for clustering generated by the algorithm (hierarchical complete-linkage clustering with max cluster size = 3) VS manual clustering using human judgement and warm feelings towards neat and clean...

Even with high correlations there is some, albeit very small, diversification advantage.

Yes, it's 90% intellectual curiosity and 10% need to construct a portfolio I could use in live trading. I'll do weightings of both clustering hierarchies and compare the weights. Perhaps you are right and there won't be much difference. I will use the new method with the preferred parameters of...

This is what I am looking at. If I run a clustering algorithm on pooled returns for rules, I get the following 9 clusters (3x3, I named the top 3): -- fast rule cluster -- 1 momentum2 1 normmom2 1 breakout10 2 momentum4 2 normmom4 3 breakout20 -- medium speed rule cluster --- 1 momentum8 1...

They are not assets per se, but rule variations. If I recall correctly, they are ewmac 2x8, normalised momentum 2x8 and breakout 10. BTW I have added a "PS" part in the previous post about the method sensitivity to the clustering algorithm.

Hello Robert, I've been looking at one particular example of correlations between 3 assets: 0.9976, 0.9417, 0.9333 (AB, AC, BC) A and B is almost 100% correlated. It makes sense that the largest weight should go to C. Additionally, the correlations are quite high and based on a lot of data...

With regards to 2 year rolling windows for correlation estimation, at first I looked at weighting the rules. I was looking at rolling 2 year correlations between different rules and their variations (ewmac, normalized momentum, breakout, doesn't matter). I found that 2 year correlation is...

By the way, one small tip is to use max distance of 0.19, instead of 0.20. Then it reproduces the original handcrafting table exactly. This is simply due to the rounding function, when, say, correlation of 0.5 turns into 0.9 when we check the "neighbour" 0.5 + 0.2 = 0.7 rounded towards 0.9. So I...

There was still a lot of time remaining before the new November article is published, I started to think how to solve the interpolation problem. I could not reproduce the method used to produce the candidate matrices and their weights (you gave a clue that it was shrinkage, but the shrinkage...

I wanted to implement handcrafting correlation candidate matrix matching method with interpolation. I ran the code to test it with correlation set [0 0.4 0], which is close to the perfect candidate match of [0 0.5 0] and got the following results: >>> cmatrix =...