@Guy

Agree.

@DR. Paul

The original question "Algo strategies based on technicals only". I think that is possible. There are many successfull systems based on technicals only. And the basic logic can be applied on many different securities. Not all of course. I do not try to "understand" the various forex pairs that I am working on. I try them. It is a trend following system and some forex pairs do not trend well.

In FX as in all other analysis, everything in effect is technical. So if one looks at the impact of interest rates on FX, the technical signals are generated from interest rate differences and applied to FX positions. When one looks at inflation, GDP growth, unemployment, consumer confidence, retail sales, again one is looking at the trends in these to deduce what positions to take.

But I agree, using one pair such as EURUSD and applying one technical signal is not going to give optimal results. It is better to use currency baskets and multiple signals to have better overall outcomes. From time to time though, one can concentrate on only one or two pairs. USDJPY comes to mind. Even there the JPY Carry would be more stable if it were diversified.

There are time though when the risks of a single currency pair have to be managed, such as when US investors buy Japanese equities or Indian buy Gold. Again multi factor models should be used.

We're talking of technicals here..Is it a concern if a strategy based on quants which works well on the Dow does not work that well on the S&P or Nasdaq?

@Ignatius, at first glance, one would answer: no, it does not matter. You have one strategy that works well and you should profit from it.

However, if someone was approching me with such a strategy, I would reject it outright as being insufficient and of little value, if any. You see, all three indices are quite highly corrolated. On that basis alone, one should expect that what works on one should also work on the other two. And if it does'nt work on the other two, then there is something wrong with number one.

If you have a trading strategy that works on only one of the three indices, then be assured it is over-fitted to the past data of that particular index. And from there, I would add that an over-fitted strategy does not transport well into its future.

So, “technically”, it does matter. I would say: it matters a lot since it would technically be doomed to fail by design and from the start.

Even if your primary objective was to achieve average performance over the long term, I would still set my goals to be profitable in all three indices as a bare minimum. If you can not beat all three indices, then you are really under performing. The conclusion to that is very simple.

The reason is there are much less working trading systems on multiple instruments or asset classes than on one single instrument. Of course it should be more robust, but with less profits on each instrument. The key here is the portfolio in both cases. I would say you have trade off between less profits and a more robust strategy. In the first case you have to change the systems from time to time and in the second case you have more work at the beginning. If you are looking for systems that work on multiple asset classes you will find some with daily bar time interval. Intraday trading systems on multiple asset classes are even more rarely .

@Andreas I can't speak for other systems out there, but in my situation there are no "trade offs", meaning my profitability does not change much from minute to minute, to daily (within. +/-5 to 8% during really challenging time. Again, there are no adjustments at all going from one time frame to anther.

I know the variance of5-8% is too high, but considering the flexibility & no adjustments, I think this is acceptable.