As mentioned in the previous post, the market of the current year has been harder for some mean-reversion (and probably trend following) strategies. I have a family of non-parametric adaptive MR strategies for baskets of equities.

I had been trading a strategy this year (on and off) that declined in growth substantially from prior years due to the changes in the market. Here is the 3 year cumulative return (note that this is not reinvested return, rather trading a fixed amount of capital):

The strategy profitability has grown much more slowly since late 2009. Here is a blowup of the YTD where you can see that it has done about 20% YTD:

The strategy appears to be more appropriate for trending markets rather than the sideways market we’ve been in. I have another MR strategy with less spectacular 3 year returns, but appears to fit this market better with fairly smooth returns YTD and more rugged returns in the higher volatility regions (2008) in the past:

A blowup of the YTD trajectory shows a 80% gain YTD:

The volatility at the beginning of the year was high with this strategy. The question becomes when to switch from one strategy to another or how to adjust allocation weights on the strategies, trading both.

### Like this:

Like Loading...

*Related*

…well, the first one looks quite familiar , sudden turn somewhere around May 09.

(http://www.quantum.meplaza.nl/onclose_sim.png)

Not surprising, this particular strategy is going long-short on a dynamic basket of equities. From what I know of your’s you are doing a single pair. You can get more amplitude with some optimisation across a portfolio.

I am selecting multiple assets and doing optimisation adaptively. But it fails to find great returns post 2010, whereas a somewhat different orientation does …

From your graph looks like you have a very steady return profile, even in recent months. What does your YTD look like?

How can you state that you made 20% YTD when you start with zero capital?

Because looking at the first equity curve, if you had started with 10$ and the same fixed amount per trade then risk of ruin would have been extra large.

And if you had started with 1000 $ to reduce the risk of ruin then you can’t state such a high YTD return with that fixed amount per trade.

This is the reason that I usually prefer to work with fixed % and start with a capital of 1. Then you can interpret the results in terms of YTD returns. Otherwise you can only say that you made xx “points”.

Do you agree or I am confused?

The equity charts I was showing are non-compounded return and not capital. The capital base could be anything, in this case was putting in ~ 1M$. So 20% over is 200K on top.