Please read the disclaimer at the bottom of this page

Sunday, September 30, 2007

New Portfolio Weights

Starting this Monday, the optimal portfolio weights (for newly initiated positions) will change. Previous optimal weights were:
Weekend Trader: 9%
Trend Plays #1: 58%
ARS: 80%

Starting next week, the new weights will be:
Weekend Trader: 32%
Trend Plays #1: 53%
ARS: 68%

Since the weights sum to 153%, leverage will be about 1.5:1.

As a result of its good recent performance, statistics for Weekend Trader have improved substantially. Because it realized its recent gains in a time when ARS and Trend Plays #1 were mostly flat, it is not surprising that the portfolio optimizing algorithm was sensitive to this and let the weight increase substantially, from 9% to 32%.

4 comments:

Eu said...

2ST Did you try back test of your systems portfolio with equal weight per system? It looks like you're in fallacy of previous statistical performance of a system. Stats is always lagging. Probably equal weight per systems portfolio is better. There isn't any recommendation from my side, just back test the suggestion ;)
Eu

Science Trader said...

Eu,
You're making a good point and I thought about equal weights across the systems. If systems show sudden large swings, like Weekend Trader, you run the risk of increasing weight at the top of the system equity curve and reducing weight at the bottom, ending up with a strategy similar to selling at the bottom and buying at the top.

However, there is one major disadvantage of having equal weights, which is that it can ruin the account (at least in theory). Updating the weights on a regular basis protects the account against a possible rotten apple.

To take an extreme example: Suppose each of the 3 system would have a 33.3% weight. If one system would show a 50% loss after a while, and the two others would each show a 25% gain, the equity in the account would be the same as you started with. However the weights then would be 16.7%, 41.7%, 41.7%. To bring everything back to equal weights you would more than triple your positions in the system that just lost half of its value, like a martingale bet. I wouldn't feel very comfortable with that.

So, regular updating the weights using historical performance is equal to "letting winning systems run, and cutting losing systems short in a gradual way".

Eu said...

(sigh) Let's make it simple. Back test your approach. You have all necessary data. So you can compare paper trading vs real trading and you can disregard your believes in the comparison.
Eu
P.S. Your beliefs are really no matter in the market. Check it, back test it. You have enough data ;)

Science Trader said...

ok, let me think about a backtest.