愚蠢的系统 (Stupid System)

This is a work in progress. I have recently updated the gains listed here based on further analysis.

This approach is ultra simple. It is completely mechanical and uses no charts, indicators, or predictions. It is based only on price action and probabilities. It is not a humongous winner, but the gains appear to be stable and consistent over many years. It seems to be applicable to most pairs.

I have run historical studies from 2003-2007 for the following pairs, with the overall gains listed:

EURUSD (805)
EURCHF (2997)
EURJPY (3435)
EURGBP (1508)
GBPUSD (404)
GBPCHF (2775)
GBPJPY (531)
USDCAD (1087)
USDCHF (-1627)
USDJPY (-7)
AUDUSD (-284)

This system is designed to take small daily profits from a basket of pairs. The individual gains are small, but accumulate over time.

The approach is based on the fact that literally 99% of the time, there is action on BOTH sides of the opening price. Only 1% of the time is the opening price also the high or low of the day. There is a very high probability that the distance from the open to the high or low of the day will be AT LEAST a certain amount.

If you calculate the average amplitude (1/2 daily high-low range) for each pair, you will find that this corresponds to the average open-high range, the average open-low range, and the average open-close range. The amplitude is a kind of basic measure of inherent volatility. That's just how much the price wiggles around (either up or down) every day.

Now, if you take 1/2 the amplitude as a profit target, you will find that the probability of hitting that target from the opening price on any given day (whether long or short) is in most cases 70% or higher.

So, here's how I did my historical studies and how the system works:

1. Use the following six pairs as an example basket: EURUSD, EURCHF, EURJPY, GBPUSD, GBPCHF, GBPJPY

2. For each pair, dump the historical data to a spreadsheet and calculate the profit target as 1/2 the average amplitude for the previous month. This means the profit target changes gradually in response to changes in the volatility of the pair.

3. Determine the "trade side." I did this in the simplest possible way ... by calculating the historical %age of up days and down days. Each pair will have an edge on one side or the other. It so happens that all of these six pairs historically have more up days than down days. That means by default that I will go LONG.

4. At the open of each trading day, go LONG in each of these pairs. Exit at the profit target or if it is not hit, at the close of the day.

5. If the previous day was a loss, do the opposite on the next day (SHORT at the open).

6. Repeat.

This example averages for the past 5 years. I have not yet worked out an optimal stop, but I will post more later on stops and drawdowns.

Other pairs/combinations will do better. However, having said all this, I would have been really happy to see it work out with every pair, but there are a few that did not, namely: USDCHF, AUDUSD, and USDJPY. There are probably others as well. I have no explanation and haven't thought about it in detail. Perhaps someone has some ideas.

I welcome your comments, criticisms, and improvements!

All the best.