Pivot Point Theory - Backtest

Hi all,

Hope your trading day was a good one, and more to come. As I live in Australia, the markets tend to be somewhat slow during my daytime hours, and so I have a little time to fiddle around with different things. Today I thought I would test out the following statement in regards to pivot points:

"If price opens the day above the daily pivot, the bias should be for long trades, if it is below, then look for shorts".

It is granted a little bit of a general statement, and I have taken a simplified interpretation of it, however one that is seen just about everywhere when pivot points are explained. It seemed an easy thing to program into Metatraders backtester, so, after downloading the latest data from Alpari since 2004, I ran a backtest with the following criteria:

 

1. If price opened above the pivot point, we enter long, going for R1, with a stop at the daily pivot. The reverse applied for short trades, where we go for S1, with a stop at the daily pivot.
2. We enter with one standard lot, and close it on the opening of the following day if it is still open.
3. Stop is placed one pip above or below the pivot depending on the trade direction (spread is taken into account).
4. Pivot points taken off the Alpari demo feed and timezone.

Attached are the results for cable, the yen and the eur/yen cross, of course there is talk about the accuracy of MT4's backtester, for a simple system like this one however, it is fairly accurate. I manually verified 6 months of the results on the charts also for each pair. Interesting indeed. What do we think?

An average of around 30-35% success rate is hardly spectacular. Like I said it is a simplified view of the above statement, but a view many beginners would certainly take. Thought it might have been of some interest.

.. and before you all ask, here is what happens if you do the opposite, long if below the pivot, short if above, with the reverse of the above rules.

Hey Akuma,

Very useful piece of information and your timing couldn't have been better. I use pivots and obviously know the rule however in my experiance, appling the rule often created resistance/confusion (mostly in my head) to good trade oppertunities as apposed to getting into bad trades. Interesting your test says it's also likely to get you into bad trades as well.

Based on this I will be making a small adjustment to my plan.

Thanks for this, Your time is greatly appreciated.

Regards
Blair

Very interesting.

I religeously plot my pivots each morning. I note that price often bonuces of the pivot, R1, R2, S1 and S2. So if these levels are at a zone where there are other reasons to trade, it is often worth the entry.

Article (Attached)

Certianly pivots as support and resistance areas have some value, what time do you take as the close and open of a day?

11pm GMT on my Meta Trader charts............

However, each morning, I check three other published sources, and take the average for each number. Uusally my charts and these published sources are within a few pips of each other.

Like any indicator level, best not to get anal about it and try and think of a particular pip level, better to consider it a zone.

The places I check are:


http://www.actionforex.com/forex_ana..._200603205734/

http://wwww.livecharts.co.uk/Members...rency_data.php

http://www.mataf.net/en/money-management.htm

Here's the thing that gets me about pivot points in Forex...the time period used to calculate them...I mean, we're talking about many variations on what the open and close bars are...some people use midnight to midnight GMT, or 8 to 8 GMT or some people use the NYSE open and close, etc etc...

I mean, if the different variations create pivot points 5 or more pips off, you can almost forget any short-term, non-trend identifying trade because you'll kill yourself trying to pick resistance/support levels.

Think, if you calculate your open/close between 5pm EST (typical US rollover), and your pivot generates 1.2860 for the pivot, but the "correct" pivot for the day is actually at 1.2875 and the price opens above your pivot and below the "correct" pivot, you're screwed. (using EITHER pivot as support/resistance OR Akuma's backtested method where if the price opens above the pivot, etc).

I asked once what time do most traders calculate pivots and I literally got 7 different response. That's when I knew pivot trading was going to be extremely difficult unless there was an identifiable trend before hand (which really reduces the usefulness of pivots unless you're SERIOUSLY short term - like 1-min chart trading short).

Pivots may be useful with confluence, but like I said, I got so turned off to pivots after experiencing the 30-35% win ratio and realizing that the "correct" pivot changed daily depending on whatever pivots the market participants chose that day.

=========================

[quote=mrmikal]
I mean, if the different variations create pivot points 5 or more pips off, you can almost forget any short-term, non-trend identifying trade because you'll kill yourself trying to pick resistance/support levels.

=========================


Don't worry about being 4/5 pips off on any indicator level. Best to think in zones rather than a actual pip value. If I have a fib calculated at 1.8800 I will consider it a zone around it.

Apart from time of calculation, different price feeds can be different by a few pips. So if I calculate a fib from a high of 1.8997, George Soros*** might be calculating from 1.8995, a few pips shorter.

***What is the main difference between me and Boy George ? I can hear your say to yourself wealth ! Not just that, he has two assistants to calculate his fibs and pivots.

[quote=hagadol]=========================

Hagadol,

I actually have pivot points plotted on my charts all the time, and you're right, I consider them more of zones...but honestly, I don't really trade them because they are highly inaccurate most of the time (even in predicting resistance/support)...For me, prices usually turn around in the middle of the points, or blows right through them. A few times (like I said, 30% or so) I find that zones act as resistance, but it's really not accurate enough for me to do any type of counter-trend positioning (either getting into opposite positions or closing positions going in the direction of the trend).

i merely mentioned the pips being off becuase sometimes we're not talking about 5, but I've seen pivots vary by 10 or 15 pips...which is actually a big difference when trying to trade daily pivots.

Akuma's backtest was solely based on taking positions when they were absolutely above or below the pivot (which was probably his own arbitrary time frame)...so any mechnical trading that used pivot points in this manner would be destroyed.

hagadol is right in treating them as zones rather than absolute levels, as an example here is the cable 30M chart of the last few days with pivot levels shown. I have circled areas, or zones where the pivots provided some insight. Of course it is always easy after the fact, but used with other methods they can help.

http://img146.imageshack.us/img146/3116/pivots3ef.gif