Model trading

As promised... I haven't had so much focus on sharing my models for you - the readers... To be honest - I don't want to share too much since I am a big believer of my models and the "edge" I believe they have... I have been doing this for four years now, spending hundreds (probably thousand) of hours developing/thinking/analyzing different trading approaches... I love this, not for the sake of the money, but for the challenge and the satisfaction it gives when I feel I have understood something a little better. I think this is an ever changing environment, and one can always learn more... The hunt for perfections never stops - that's why I have developed more than 100 models, but only a few feels good enough to trade live in my view. One who has made the cut is the one I am trading on in this challenge.

I will try to describe a little bit better what is behind the model:

The model, let us call it the FX-macro, is based on the relative economic strength between different countries... The rationale are the following, if a country (A) has a strong economic development (increasing GDP, declining unemployment...) vs. a country (B) with weak development (declining GDP, increasing unemployment) then investors should be more interested to invest in A than in B. Since the FX-market is all about demand and supply, then the currency corresponding to country A should be in demand (inflow of capital) while country B's currency should be well supplied (outflow) - implying that the currency A/B should increase in value... What kind of indicator do I use for the economic development in different countries? I use assets from outside the FX-world - mostly rates. This works for me, but somelse may be better off using equities or something else... So, to sum up, if a country is expanding relative to other countries, than it is quite likely that I will be a buyer of the corresponding currency.

I just wanted to include two graphs - highlighting the usual problems as a model trader...

I currently validates an intraday momentum-model which I have developed a couple of years ago... The first graph shows the result when I start in the morning while the other in the afternoon... The difference in result is huge! This seems a bit fishy... :) The first graph is beautiful... Something is wrong, I haven't found out yet (will take some time to go through all code), but this is a usual situation when developing - "learning by burning"...