The Luxor System

I have recently been experimenting with the basic Luxor system described in the book Trading Systems by Emilio Tomasini and Urban Jaekle. I know there is a previous thread on here discussing this system but I'd like to resurrect the topic as there were no backtest results shown in that thread.

The system is a very simple moving average cross for entries but with a breakout filter added, where once a cross takes place trades are only entered if the price breaks the high for longs, or the low for shorts, of the previous bar. The idea is that this provides confirmation of the price move. The results put forward in the book are surprisingly good for such a simple strategy. The author presents some results from TradeStation on GBPUSD over 2002-2008 showing a net profit of $66k from fixed 1 lot trades and a decent looking equity curve.

I don't use TradeStation, I use MetaTrader 4. So I've coded up this strategy in MQL as described and can't get backtest results anywhere near what is put forward in the book. On GBPUSD (FXCM data), over the same data period and timeframe the equity curve is all over the place and results in a loss. I've attached a screenshot of my results below.

Has anybody looked into this system and had any success? and more importantly has anyone actually been able to replicate the results shown in the book?

I'm not sure if it's due to a difference in the data used, something I've misinterpreted in my coding, or if the strategy just doesn't work.

TradeStation users - would anyone be willing to test the Easy Language code provided in the book (attached below) and see if the results come out the same as shown in the book?... (screenshot attached).

For me having a system where I could have 16 consecutive losses tell me it's not worth trading this way

Why? Consecutive losses tell you nothing about a system. When you have the right R:R then such a system could be profitable. The other extreme is martingale then you need only 1 losing trade to blow an account.

But don't let us highjack the thread with that.

So your saying that if you start off with 16 straight losing trades you want to continue? really I guess your not a full-time trader. If I have 3 in a row I go and reevaluate haven't done that for a long time good luck trading this

so you are a full-time trader? Then think about your understanding of risk management. When you evaluate a system with consecutive losers you have no understanding about risk management. You would not trade a system where you have for example 20 losses a 0.1% and then a winner of more than 2%? Do your homework! If you can't withstand that psychological that's another page, but then it's your problem and not the problem of the strategy.

But as most "amateur full-time traders" you want high winrates to impress other amateurs, aren't you?

Wow thanks for the spanking yes I trade fulltime for a living I day trade the e mini nasdac futures market. Amateur no but as I see your a dreamer as soon as you mention the word Martindale. Good luck see you in the funny pages risk management hahaha 16 losses in a row are unexceptiple

I should point out that the strategy logic and results shown above are not the final system provided in the book. This is presented as the initial concept and then the author goes on to discuss adding additional filters such as a time of day filter, along with some stoploss strategies, and some optimisation of the parameters to improve the system. He actually manages to double the profit, get the profit factor up close to 1.5 and reduce the drawdown (results attached).

However none of it is much use as I can't get the initial concept to run a similar profit to what they have in the book let alone any adjusted version.

Hoping we have some TradeStation users on here who might be willing to backtest the Easy Language code to compare results?...

......

On the consecutive losers debate I have to agree with JensItZig. For my automated strategies I watch the drawdown $ and %, as long as they are within my tolerance I'm not concerned. These kind of trend following strategies do tend to have a large number of losers when the market is sideways but make up for it when they catch a trend. I do appreciate it would be difficult to trade this type of strategy manually though.

When I read strategy test report, I often check largest loss trade -26.84 and average profit trade +8.2.
If largest loss trade is great than average profit trade, then, this trading system (EA) is bad one on long run.
It skews toward loss.

This is my only criteria to assess if a trading system (or EA) would win in forward living trading.
This is a top confidential tricky in living trading.

I see you don't even understand what i wrote. You wrote that you rethink a strategy even after 3 consecutive loss. My example of martingale is only to show YOU that even a system where only 1 loss is needed to blow an account is rubbish. The only thing you should understand (trading for living or not) that it's all about risk management and not consecutive losses.

And how dreaming is linked with correct risk managemnt and correct evaluation of systems will be your secret. Do the math and statistics!

Und auch noch aus der Schweiz. Ist euer Bildungssystem so schlecht? Wow.

hi,
I am just reading this same section in the book and I am also an mt4 user, and before trying to invest the time and effort to translate the code to MQL I wanted to ask were you able to replicate their results of the initial part (where they leave the exits out and only take the entries from the system)?
Thanks

I'd respectfully suggest evaluating the system - all systems - based on expectancy.

"Quick & dirty" - but nevertheless useful - formula is "Average trade / Average Losing Trade".

An extraordinary statement.

I can only assume you meant to say that within reason consecutive losses considered in isolation don't tell the whole story. That's true, but it's very significantly in three places from what you said.



That's true, but firstly this one doesn't, and secondly you need to consider three other things, also:

1. the adverse potential on position sizing

2. the reality that although a system with so many consecutive losers may be profitable, it can also be highly stressful and difficult to trade

3. the reality that while long consecutive losing runs are very rare, long losing patches are very common, and a system that has 16 consecutive losers is pretty likely, at some point, to put you financially into an overall position equivalent to 50 consecutive losers, even without any particularly long losing runs.

I can only assume you meant to say that it's very significantly different from what he said.

But an otherwise excellent post, and it needed to be made, too.

Please read the whole thread and not judge single posts!

There are statements that systems or strategies can be judged by there consecutive looses. Thats rubbish and thats all i try to explain. I never wrote about the problems you can face by trading a strategy with high R:R ratio and some longer loosing streaks. If you can live with the pressure then it's not your tradingstyle (i personal wouldn't also trade those).

hey, did you manage to fix the code? i was interested in this strategy but didnt actually got to coding but was curious if it still worked

Hi parkzo, did you manage to solve the problem with the luxor system?