Why Backtesting does not work: An Overview - page 2

New comment
 
MDET:
1. If strategy doesn't perform on backtest it most probably won't perform in real trading;

2. Past doesn't repeat itself as much as somebody would like it. Because of that backtest is even better as Demo trading. You can't loose or win, but you can at least check pount 1. Weel, back test is even less emotional as Demo, but you can backtest years in seconds.

3. Backtest is very helpfool exploring strategies, how do they perform in different conditions, what is worst and best results etc.

4. Can you use bactest as proof for future wins? No. Backtest can just give you idea of probabilities. Thats it.

I feel the differences between back testing and live trading lies in the fact that in backtesting there are NO open trades. I have witnessed first hand how you are more likely to hit targets for SL and TP if you are trading live without having them set with your broker. Simple Answer BROKER MANIPULATION. In backtesting the broker can not move the market to "Take Out" your stops.

 

BT and conventional trading--

why BT will never work >> good post ... agreed too-

i try and keep it simple [after few years of market research tho]

why BT will never work : trying and crack the market (if one wants to get out of the RR, martingal/anti-, and other probabilistic ways) requires a bit more than using indiz the way they come (as this wont work unless several other variables are added) and thinking BT can help you decipher the market in the LT range is the mistake of a life time and pure misunderstanding of the very nature of the markets.

BT will not REPLACE the knowledge a young trader HAS TO acquire -- that s a given. (and i dont see HOW BT can help, as some people suggested ; as it is like sailing and trying to analyse and give meaning to every wave you are riding ... they are all different ALTHO they have something in common ... more on this)

To try and understand the market, one has to get closer to what is holding the general rules of Universal laws ... so good luck if you try and "scalp" without a clear plan.

This is the most ruthless business ever, being armed with BT only will get you nowhere ... even with large account.

You need at least a well spread system, with several variables and set-ups...

anything less than this will bring you back into the 95%-clan.

Then in the LT (long term) you need to acquire knowledge other than learning formulas by heart or mastering R/R scenarii.

...for a little hint : direct yourself towards Strange Attractors.

and stay off EAs, analysts retentive sh**, news trading and beware of your broker(s).

g/lk to all

exni

 

BT doesn't work, but it CAN be helpful.

 

The article's author expect we pass 30 to 50 years sit and looking the monitor to understand price behavior, patterns, and indicators reaction.

200 years after the first market here it comes this guys to tell us: backtesting does not work.

Not a practical approach. Right?

So, according to this guy:

Elliot work is a waste of time.

Pattern studies is is a waste of time

Pring work is a waste of time.

Merryl work is a waste of time

Gartley, Pasavento, and thousands books, magazines and articles are a just waste of time.

Interesting approach

 

Hedgers

Hedgers :

Some of the biggest clients of these banks are businesses that deal with international transactions. Whether a business is selling to an international client or buying from an international supplier, it will need to deal with the volatility of fluctuating currencies.

If there is one thing that management (and shareholders) detests, it is uncertainty. Having to deal with foreign-exchange risk is a big problem for many multinationals. For example, suppose that a German company orders some equipment from a Japanese manufacturer to be paid in yen one year from now. Since the exchange rate can fluctuate wildly over an entire year, the German company has no way of knowing whether it will end up paying more Euros at the time of delivery.

One choice that a business can make to reduce the uncertainty of foreign-exchange risk is to go into the spot market and make an immediate transaction for the foreign currency that they need.

Unfortunately, businesses may not have enough cash on hand to make spot transactions or may not want to hold massive amounts of foreign currency for long periods of time. Therefore, businesses quite frequently employ hedging strategies in order to lock in a specific exchange rate for the future or to remove all sources of exchange-rate risk for that transaction.

For example, if a European company wants to import steel from the U.S., it would have to pay in U.S. dollars. If the price of the euro falls against the dollar before payment is made, the European company will realize a financial loss. As such, it could enter into a contract that locked in the current exchange rate to eliminate the risk of dealing in U.S. dollars. These contracts could be either forwards or futures contracts.

Forex Market

 

Ok, this article makes a good point, but what if you only backtest half of your data and optimize it and then backtest the other half as a test?

I think someone already mentioned curve-fitting but I just wanted to ask this anyway.

 

Yet the logic of backtesting depends precisely on the validity of that it is possible to make such predictions. Adherents of backtesting methods believe that if a strategy can be optimized on the basis historic data in such a way that a perfect match is obtained, it can be used to predict future events as well. This belief goes against the most basic rule in trading, proven by countless examples, that past performance is no guide to future results. In addition, if we presume, as discussed in the previous section, that technical tools cannot predict future market action, what is the point of testing how good the predictions of any technical strategy could be based on a comparison of past and future data?

How to test a technical strategy?

We must admit that a technical strategy can only be tested in live trading. The same problem that make demo trading useless as an introductory phase to live trading ensure that the results generated by backtesting methods are more or less irrelevant in establishing the credibility of a trading strategy. To test your strategies, test them in the difficult conditions of the market, because testing a strategy involves testing yourself too. The methods you employ are strongly influenced by your risk tolerance, emotional resilience, patience, and timeframe for success. Backtesting cannot account for these factors, and it is better to avoid it as a result.

It is also important and useful to do away with the notion that there exist come perfect strategies that can be discovered through backtesting methods. There is no question that such strategies do not exist: the discovery of such a breakthrough would perhaps be tied to a mathematical breakthrough, and it is improbable that such an achievement will be the result of the endeavors online marketers and peddlers.

Conclusion

The only value of backtesting is for educational and entertainment purposes. You can use it to familiarize yourself with the basic meaning and use of various technical indicators, and to become accustomed with the mechanics of creating and applying a technical strategy, but it’s not possible to make use of backtesting for the purpose of making predictions, or establishing the reliability of a trading method. Backtesting is useful to marketers and peddlers because they will use anything that promotes the value of their products, be it fair or foul. But traders need more than empty talk, and advertisement for profitability, and that’s why any claims must be made subject to rigorous testing and questioning. It is clear that backtesting fails in this respect.

If you desire to test your strategies, yet are worried about the risks involved, making use of a mini account is always the best option. There are brokers that allow no leverage trading with as little as $5 in today’s forex market, and such an account could be used for a very long time with only negligible risk for the testing of a potentially worthwhile strategy.

By Forex Traders

 

This is a very good article. Backtesting is the key that will show a trader that the disciplined strategic trading is better than intuitive only because emotions stay out with trading with a technical strategy.

 

Even twenty years later there are still folks that swear that backtesting simply doesn't work.


And they're right!  If you don't do it well it doesn't work very well.

 
Whoa... truth hurts ;)
===
The only value of backtesting is for educational and entertainment purposes. 
===
12
New comment