As retail traders, should we stop searching for holy grails and embrace mediocrity?

[KingKaivar]

This is a good mindset to adapt.

What kind of systems are you trading? I’d be interested to learn a bit more about it.

ATM I’m backtesting an ATR-based system. Setup ATR(14), and if the daily range is less than 50% ATR(14), enter on the break of the daily high or low in the next day. Your stop is placed on the opposite high or low of the break. Very simple. My exit is when I hit 2:1 R:R. Expectancy = around 0.3R, from backtests on the EURUSD and USDJPY from 2001 to mid-2012. Sample size = 300 so far. I manually backtest my systems so it takes awhile but the good thing is that you always develop new ideas as you go along.

I’ll go through the AUDUSD, USDCAD, GBPJPY and gold during this week and eliminate duplicated / highly-correlated trades at the end, but I don’t see why the system should fail. If you trade the majors + commodities, you’ll probably find 3-4 trades per month.

There’s some other systems I’ve designed and backtested, but I need to go to work. I’ll post some concepts / ideas in a future post.

[MeiHua]

My trading methodology is a little different. Rather than spending years “perfecting” 1 trading system. I like to trade several at once, assuming they all have some sort of an edge. Typically, I look at the profit factor of each trading system and if they are above a minimum standard, I’ll usually run it. The point is that I’m looking for frequency over quality. I guess in my case, quantity over quality is true! My risk is small, 0.1% - 0.25% for the most part, but I’m taking 100-200 trades a week, with a statistical edge. You can say that I’m essentially speeding up the “law of large numbers” process!

But I do want to say that I am mostly automated. And that, I would agree with you, that as a manual, discretionary trader, my methodology probably will not work.

This really is key in my trading as well.

What is your minimum criteria for your trading systems? Which are your top fitness functions you check? Also for an individual system how many trades per day/week. The only reason i ask is because i am looking at creating an intraday algo. I dont want to derail this thread though but would like input.

---

In regards to embracing medicrety, it depends on how you execute your edge. For some its larger big R:R trades that are infrequent, for others its a big round of lots of piranha bites taken out of the market. the key is here, they all have edges. You can turn any edge into something where people would be astonished at the result. How you came up with the edge, either through experience, whatever does not matter. Also execution, discretionary/manual or automated does not matter. Its all about expected value * N . the higher your expected value the less N you need to meet some goal, if your expected value is lower you need more N. either way both people make money.

[ClarkFX]

ATM I’m backtesting an ATR-based system. Setup ATR(14), and if the daily range is less than 50% ATR(14), enter on the break of the daily high or low in the next day. Your stop is placed on the opposite high or low of the break. Very simple. My exit is when I hit 2:1 R:R. Expectancy = around 0.3R, from backtests on the EURUSD and USDJPY from 2001 to mid-2012. Sample size = 300 so far. I manually backtest my systems so it takes awhile but the good thing is that you always develop new ideas as you go along.

I’ll go through the AUDUSD, USDCAD, GBPJPY and gold during this week and eliminate duplicated / highly-correlated trades at the end, but I don’t see why the system should fail. If you trade the majors + commodities, you’ll probably find 3-4 trades per month.

There’s some other systems I’ve designed and backtested, but I need to go to work. I’ll post some concepts / ideas in a future post.

Hey Kaivar,

That’s great! Looks like you have a good idea of what you want to be doing, wish you the best of luck! If you need helping getting some accurate backtesting results, with some actual decent tick data not from MT4, let me know. I’d be happy to code something up and help you out!

Clark

[ClarkFX]

What is your minimum criteria for your trading systems? Which are your top fitness functions you check? Also for an individual system how many trades per day/week. The only reason i ask is because i am looking at creating an intraday algo. I dont want to derail this thread though but would like input.

---

In regards to embracing medicrety, it depends on how you execute your edge. For some its larger big R:R trades that are infrequent, for others its a big round of lots of piranha bites taken out of the market. the key is here, they all have edges. You can turn any edge into something where people would be astonished at the result. How you came up with the edge, either through experience, whatever does not matter. Also execution, discretionary/manual or automated does not matter. Its all about expected value * N . the higher your expected value the less N you need to meet some goal, if your expected value is lower you need more N. either way both people make money.

In terms of analyzing trading systems, I don’t have a concrete procedure for weeding out systems. It’s more about looking at the overall picture for me. Of course some key stats that stick out is profit factor, Sharpe, Sortino, drawdown, frequency, etc. As far as how often each system trades, sometimes 10 trades a day, sometimes 2 a month. I like to have a portfolio of non-correlated trading systems as well as trading systems of all types (scalping, trend following, mean-reversion, etc).

[KingKaivar]

In terms of analyzing trading systems, I don’t have a concrete procedure for weeding out systems. It’s more about looking at the overall picture for me. Of course some key stats that stick out is profit factor, Sharpe, Sortino, drawdown, frequency, etc. As far as how often each system trades, sometimes 10 trades a day, sometimes 2 a month. I like to have a portfolio of non-correlated trading systems as well as trading systems of all types (scalping, trend following, mean-reversion, etc).

Hi Clark. What resources did you use to study and develop your style of trading? I’m especially interested in how you manage to use different systems simultaneously, and what you do to test for correlation? My current correlation test is to check if the systems I’m testing provide an entry signal at the same time. I’m wondering when does a correlation value become “too high”?

[AthenaFx]

I’ve probably designed at least a dozen marginally profitable trading systems (0.2 to 0.4R expectancy), but thus far have failed to come up with a high-expectancy system (>1R) that can compare with professional traders.

Then I asked myself, why fight using my weakness? Why not deploy all these mediocre systems and let frequency of trades compensate for my low expectancy? The more systems you use, the more trading opportunities you will find.

Obviously these systems should be non-correlated as possible. But as a retail trader, I think this is one way we can shift the odds in our favour against the professionals and institutions. Thoughts?

Retail traders are mostly “momentum” traders.

There is nothing wrong with that but they have to open their mind and see the market in a different way.

This is a required condition to complete a crucial step in the development of trading skills.

[FXAssassin]

Retail traders are mostly “momentum” traders.

There is nothing wrong with that but they have to open their mind and see the market in a different way.

This is a required condition to complete a crucial step in the development of trading skills.

Athena…can you clarify what you mean by retailer traders being primarily comprised of momentum traders?

Also, what way do they have to adjust their thinking to make it fall in line with being a momentum trader?

[Fxmall]

Naked trading…now that’s something we hear and tend to ignore when searching for the “holy grail”…and yes, it is always a good idea to never force the issue in a trillion$$ a day market…