Article / 23 November 2016 at 14:00 GMT

Walking our way into an average strategy

Hypothesis Testing
United Kingdom
  • Similar to pedestrians, markets also move in different directions for short periods
  • For this project we consider strategies that involve random walk scenarios
  • Profit expectation is strictly the mean reversion
  • We're looking only at close price differences and seeking 50% of that difference

i It  looks random but those Tokyo folks probably know exactly where
they are going, allowing for the odd meander or two. Photo: iStock


By fxtime*

Let's first state that I will not post my trade strategies from earlier in the year as I suspect we all have the gist of standard deviation, Bayesian and end-of-day postings for whatever liquid market that I and other more learned posters have made on the site.

Next consider random walk scenarios. When we walk from one shop to another we walk directionally with the expectation that we will successfully achieve this simple task. However, we are also aware of fellow pedestrians and will weave our way through other people's random meanderings. The market is no different. Expect moves in either direction for a minor period of time while the market makes its current trend move.

For this project I am using cash market spot prices of the cable (GBPUSD) as then we have a clean database to work from and it's easy to backtest for years where as backtesting futures has its problems. The cash market will not have expiry days that prevent trading nor will we have to consider rollover points and subsequent liquidity traps.

Stops are preset for a zero-sum basis. In other words, any trades you make over say a three-month period will at worse bring you to the same balance figure. Therefore you are no worse off. Consequently if we apply skew, which tightens the stop to below the zero sum value, you know that over time your account will consistently grow. 

This one paragraph should be the concept basis for all your trades. Look at your account balance now and then go back through your account statements for three months and you should be wealthier or at the very worst just the same balance.

There is no trade size being quantified for this trade set up as we only want to build a scaleable strategy so assume our success/fail outcomes will be displayed in real-time points and not in any currency value.

So let's begin.

The most liquid time for this market is 0800-2100 GMT so we will only consider the most liquid timeframe for ease of trading. So far using the close price of each 4-hour candle we seek merely the mean reversion of the current and immediately prior candle. 

Profit expectation is strictly the mean reversion. Remember we are looking only at the close price differences and seeking 50% of that difference. Profit target level is defined and limited to this level. Never try and exceed this as that is an untested theorem, which incurs additional risk. We are looking at steady income only.

Next we need to look at what the pre-market ranges are. After all we know Asia has been influencing this market before we even wake up so we need a guidance to start from.

Pre-market ranges I utilise are:

  • FTSE100: 0400-0800
  • EURUSD: 0400-0800
  • Cable: 0400-0800

Let's just use the close price of the 4-hour candle of the above range and compare to the prior 4-hour range close price. Remember we are only comparing close price to close price. Nothing else. The usual market event is to move to the mean of these two reference points. However, the drawback is that we do not know when this will occur.

What we do know, however, is typically how long it takes for the market to complete a mean reversion move described above for 50% of the time or 70% or using a basic distribution curve the optimum number of times you can expect a successful outcome.

Yes we are describing our old friend the zero sum again. We need to find the best trade success to ensure we don't erode our account balance over a period of time – say three months – due to a rogue spike in the marketplace.
k



An  equal and exact opposite effect. Photo: iStock

Now we know where we should be looking for trade success and typically what duration of trade we can expect before booking a profit on balance. We also need to consider if any other signal appears subsequently and should we enter a trade for the next signal.

The early pre-market range is invariably small for the potential profits available so we need to see where our chosen market wanders after the opening bell using the following additional 4-hour close price ranges.

  • FTSE100 cash market 8-12
  • SP500 cash market 8-12/12-16/16-20
  • EURUSD 4-8/8-12/12-16/16-20/20-21
  • Cable 4-8/8-12/12-16/16-20/20-21
  • USDCAD 8-12/12-16/16-20

This also works, for what it's worth, on Dax, USDJPY and AUDUSD – in fact on most liquid marketplaces. You just need to look for the optimum pre-market 4-hour range.

In the case of the FTSE100 at 1200 (noon) when the next 4-hour candle completes we can compare the close price at 0800 and 1200 and enter a contra mean reversion trade with a 86% probability of success within three subsequent trading (4-hour) candles. Three subsequent candles seems very high but the data reflects the maximum period and it is common to see a move complete in four hours. 

The drawback of these trades is obvious: we seek ONLY 50% of the two close price values and we are exposed to unknown adverse market moves. Likewise subsequent candles may well give an opposing signal, which you should take as we are trading from a probability model and each 4-hour close price range is an independent trade strategy to take that is NOT conditional on the prior 4-hour candles' movement.

Last week I posted two such trades on the site. Both proved successful. The success rate on one is currently 22 successes and one failure; the other was 87% success. To maintain fiscal control there are a number of trade stop scenarios you can use – my favourite is a monetary stop. 

Select a fixed stop that you feel is the maximum loss you would be comfortable with. Also trade small and often. Small trade sizes inflict little pain for drawdown or stops but the frequency of these trades rapidly accrue revenue for you.

Stops are paramount as events like Donald Trump's win or Brexit would seriously damage your account. These two events were obvious trades to avoid likewise the Federal Open Market Committee should be avoided as the market will be illiquid.

Attached are two charts. The first is the euro spot market using an algo function of mine. It made 25 trades and 23 were successful over the last week.

Total gain: 268 pips
Total losses: 74 pips 

g

Source: SaxoTraderGO

This is the cable chart for the same period.

It shows pretty much the same outcome except there were three duff trades where the 50% reversion point failed, but two of those trades still went into profit and at worse you closed out at break-even.

f

Source: SaxoTraderGO

Even if I was trading on a spreadbet account using the minimum £1 per point stake size, my account would be in profit in excess of £500 over one week. Perhaps being average is good?

As usual I will post real-time updates and charts. Hopefully if nothing else it makes readers look slightly differently at their charts and then look at hourly charts too for these scenarios.

– Edited by Gayle Bryant

*fxtime is an alias
23 November
fxtime fxtime
Update for this week...ftse chart
23 November
fxtime fxtime
Update for this week...DAX chart
Note the most recent signal reached within 6pts of target profit. There was plenty of opportunity to manage the trade and earn revenue. But the DAX has a quirk whereby it often completes a missed fill the subsequent day !
23 November
Estuardorlemus Estuardorlemus
Great article fxtime. Really helpful info.
23 November
fxtime fxtime
You are welcome :-) Lets see if the DAX follows its habit of retracing to make a mean reversion !? I should really describe how to trade binaries on the above but not sure if readers use binaries at all.
23 November
Estuardorlemus Estuardorlemus
Any information is welcome ;)
23 November
goldfinger goldfinger
did you trade for LTCM?
23 November
Scoot Scoot
fxtime, thank you for the note. I seem to be missing a couple of things however.. 1. how do you select after what periods to trade.. as not all candles on your charts have "trade" triangles. 2. can you go a bit more into how you select stops and if its a function of max gain on each trade. Thanks!
23 November
fxtime fxtime
Some trades run out of time for opening...eg after 8pm if a tarde isn't already opened then the trade order will lapse and await a new time frame signal at 8am GMT. As for profits and stops.....the simplest method is to use small stakes .....so if you normally trade £10pp with a 50pt stop then trade at £2pp with the same monetary value stop. if the trade becomes more extreme you can scale to say a total of say £3pp but your monetary stop MUST remain the same and adjust your stop accordingly. Do not increase fiscal risk. Alternatively use binaries as returns are leveraged and risk/stop loss is fixed.
23 November
fxtime fxtime
Goldfinger I do love the LTCM trade setups...they over leveraged though but their strategies for option plays were ahead of their time imho.
23 November
Scoot Scoot
"Some trades run out of time for opening...eg after 8pm if a tarde isn't already opened then the trade order will lapse and await a new time frame signal at 8am GMT." sorry what i am missing is what the "signal" is to open a trade? is it not simply the close of each 4hr period? Sorry i am missing something
23 November
Scoot Scoot
i am also wondering if the s/l may be a function of the expected profit, like 2x of your target in pips.. have you backtested something like that to see if you do not get stoped out of potentially profitable trades?
23 November
fxtime fxtime
In the article it should show the time frames used for opening trades. Normally I trade upto 8pm GMT....my algo's can run through the night but this is for manual trades. No new trades will be opened at 8pm onwards but i might have trades running from prior candles and I will leave them to run. Normal manual trading as per the article starts at 8am (the 4am-8am OHLC close value) .
My take profit is obviously limited to the 50% target level but using binaries enhance results...if price moves further away from my expected level then it presents a better b/e providing I keep monetary values fixed.
23 November
fxtime fxtime
These trade structures are basic scenarios that permit adaptation and over time other strategies will overlay and tighten the tarde set up. Over the entire year that is what i have been posting.....different strategies/indicators/filters that permit a trade setup.
23 November
Scoot Scoot
yep, all very interesting thank you. Naturally, looks like it works well for sideways markets.. just looking at usdjpy this week this wouldn't have worked well from the looks of it
23 November
fxtime fxtime
That is why I use binaries....costs/risk is defined and returns rapidly accrue. The major flaw of all mean reversion trades is a run away market and the limitation of profit. That said there things that can be used as a filter to know when to fade the market or not and I will post those in the live updates. I did have one such filter occur today but it happened before the article was posted so lets await a real live event and i can post as these occurrences appear.
23 November
Trader Mick 2 Trader Mick 2
I think I understand this but just to confirm (please bear with me) - If you have two close prices of say 100 & 112 the mean price is 106. At the close of the 112 bar a trade can be placed with a target of 109 (50% mean reversion)? Looking at your charts this seems to be ok but I just want to confirm this is the case. Thanks again for your articles and advise, I have been working on trading like this for quite some time and you have set me straight.
23 November
fxtime fxtime
Spot on Mick. Often what happens is that the market will exceed the 4hr range before reverting to mean so you can time the trade entry to your preference but the core concept you have :-)
23 November
fxtime fxtime
Attached is the update from the suggested long trade on the DAX...the chart is in the commentary. I have stripped out any other trades that earned on the DAX and post only the raw results to show the LIVE outcome.
23 November
fxtime fxtime
Here is a failed trade from yesterday...the cable which has now worked LOL
https://www.tradingfloor.com/posts/i-wonder-if-the-spot-cable-market-recovers-back-upto-124630-today-i-suspect-fwiw-it-will-8250204
chart shows the explanation etc.
23 November
fxtime fxtime
As for request about back testing...all my trades and articles this year have been back tested and operate on a skewed zero sum basis to ensure underlying account growth.
https://www.tradingfloor.com/posts/applying-zero-sum-games-to-a-trading-strategy-8059575
23 November
kom75 kom75
Hi fxtime. Thanks for another lesson. As for Trader Mick 2 example. If difference between two closing price is 12 (100 & 112), wouldn't be 50% if mean reversion at 106?
23 November
kom75 kom75
If=of:)
23 November
fxtime fxtime
Yes it would LOL...many thanks for correcting.....clearly I am cracking up LOL Thanks mate :-)
23 November
kom75 kom75
You're welcome. Interesting day btw:)
23 November
Trader Mick 2 Trader Mick 2
oops...I mis-read. Its been a long day looking at charts. Thanks for clearing it up Kom75.
23 November
kom75 kom75
No problem. Always happy to help:)
23 November
fxtime fxtime
i always like your ability to understate Kom......It has been a great day....the usdcad caught me off guard but the above along with teh usual trade set ups and the ftse drop was superb so my guess it has been for you too :-)
23 November
kom75 kom75
Not so much. I recover now from short usdjpy trade. Closed to early, but I'm short again. I didnt trade breakout on real acc, but on demo was very good, especially short eurusd. Overall if I would trade all algos on real acc instead of demo, day could be fantastic:)
23 November
kom75 kom75
Commentary above regarding my algos can be very helpfull for people who just started to follow fxtime. All algos are based on one or more fxtime's strategies and they really works:)
23 November
fxtime fxtime
Very kind of you to give feedback Kom :-)
23 November
kom75 kom75
Only tell the truth:)
23 November
AlexF AlexF
Hi FXtime are you still there.....?
23 November
AlexF AlexF
Ill post a few question using your EURUSD chart I also guess I understand part of it but still...
23 November
AlexF AlexF
Can you first look at this one, do i get it right. ?
23 November
AlexF AlexF
For me these is something wrong in my interpratation since you have several red candle one after each other with a successfull trade Ill show in next one.
23 November
AlexF AlexF
HYG
23 November
fxtime fxtime
First query...chart attached.
23 November
AlexF AlexF
1.242-1.2311/2 = 53 pips more or less so ok but why you say two candle later ? The next white candle is up 50 pips
23 November
fxtime fxtime
yeah but the profit target was reached the next candle.
23 November
fxtime fxtime
query on the euro chart you asked....enlarged screen for that day for you
23 November
AlexF AlexF
ha ok ....
23 November
fxtime fxtime
These trade set ups do not know what the next candle will be for direction all I am doing is playing probability curves as to whether the market will retrace 50% of the difference between the two close prices and what time frame gives the optimum return. As you can see some require 24hrs but the optimum use of funds and stops etc is the the subsequent three candles.
23 November
fxtime fxtime
Thanks for looking/commenting mate and have a good evening :-)
23 November
AlexF AlexF
Got it thanks ! FX you do this manually ? in the time frame you mentionned everyday or you have an automoated way of doing it ?
23 November
AlexF AlexF
BTW if it is not asking too much can you just tell us your SL on the short position on this chart as an example
23 November
fxtime fxtime
I operate automated algo's and do manual trades too ! The short on that particular trade was an algo so it was an out of hours 100pips once the trade moves in the desired direction the algo shifts to use the prior candle as my trail stop.
23 November
fxtime fxtime
Probability trades do not apply classic 3:1 profit ratio trading functions...they merely look for a rolling data set that gives a probable success rate over a predescribed value. When you drive your car 84% of the time you remain in your lane...the rest is for parking, turning, overtaking etc etc. You have a very high confidence value that you will stay in your lane safely and it is much the same for these. Confidence values are close 1.00 and probability values are high 80's percentile. The higher the percentile though the less profits overall you gain in any one trade but the odds are with you favourably.
23 November
AlexF AlexF
FX are you training people as well ? Ill meet you in UK anytime just say the prize : )
23 November
fxtime fxtime
I do train people LOL....but usually as a group. I will let you know when the next one occurs in London and you can join in.
23 November
AlexF AlexF
That;s a deal please dont forget and if you have any contact details please post. I did learn quite a lote tonight after a long day at work that my way to relax
23 November
AlexF AlexF
thanks a lote
23 November
fxtime fxtime
There should be one in the New year...i will let you know and have tagged your info for sending info. Have a good evening mate :-)
23 November
AlexF AlexF
Excellent thanks
24 November
ajm ajm
Hi FXtime you can count me in as well on next course in UK in the new year, i need to learn/understand more about algorithms and binaries..and which programs/tools to use
24 November
fxtime fxtime
OK thx :-)
25 November
Rogan Rogan
Excellent article, you are a cornucopia of ideas...I am struggling to keep up LOL
25 November
fxtime fxtime
LOL....as long as they are useful !
Have a good weekend mate as volatility should return next week :-)
25 November
Rogan Rogan
Very useful indeed...talking of volatility , the Italian referendum should provide a few opportunities!
Enjoy the weekend mate
28 November
Feders Feders
Hello fx, reading this articule several times to get it all. Very interesting as always with whatever you produce.
But still one thing I did not understand is the point where you say that "Next we need to look at what the pre-market ranges are", and you mention several markets. Is not clear for me what do you use these ranges for? What information do you get from them useful for this strategy?. Can you please going deeper in this point?
Other thing is the stop loss point. I have a small account and so my monetary stop loss is low, and I was applying this strategy randomly with my monetary SL and in almost all the cases the SL would be reached before the profit limit. So to use this idea I have to use a larger SL.
Thanks FX keep forming us!
28 November
fxtime fxtime
I can see your point and the phrase about pre-market really relates to this weeks article and all i ws doing was building a base for next week so the articles easilly overlap. IF you use the premarket range before a mainstay index opens you will find the range between 4am and 8am say for the FTSE100 is small and unlikely to be worth trading unless the subsequent candle moves away from the 50% level permitting you to have a better trade entry. But that will become apparent with this weeks article. These trade scenarios require patience and you are correct that stops cannot be applied in the usual sense for points away from trade entry. That is why I suggest as a basic scenario a monetary stop value so if you are willing to risk £100 stop make your stake £1pp trade very small initially and build your trading reserves. The stop is very wide as we are merely seeking probability and you would have seen the trades work reliably.
28 November
fxtime fxtime
These trades are scale-able either by options or binaries for enhanced returns or you can add to your small stake position from £1pp to £1.50pp but KEEP the same monetary stop value of £100 at risk only. Thus your stop tightens but you gain on revenue etc....the point is too benefit from probability scenarios and the distribution curve. Trade small at all times.
28 November
Emil Filip Emil Filip
Hi fx! Long time since i didn't talked to you but still reading your posts which as usual are great. Please add me on your list for your next training group. Thanks.
29 November
fxtime fxtime
OK thanks :-)
22 February
trendy trendy
As far as I see, the trader go offer different 4h intervals: 6-10 am, 10-14, 14-18, etc; can be set as 4-8 am, 8-12, 12-16?
22 February
trendy trendy
Sorry, it was my local hour: set to gmt...
22 February
trendy trendy
As observed in your two charts from the article, the failing trades was after two very long candles. Maybe should find some filters, to sort out these kind of trades; ATR wouldn't be helpful?
22 February
fxtime fxtime
Don't know what market you were trading but look at the spot cable for the last say 2weeks set at 4hrs on your charting. Seriously reliable profits there. If you were expecting mean reversions to be quick well they aren't but can be good revenue earners overall especially the cable the usdcad too for that matter. But they take time.
22 February
fxtime fxtime
@trendy the cable chart for you fwiw.....look carefully they ALL worked....including todays.
22 February
trendy trendy
Maybe I don't understand the strategy... but I see a long bullish candle in 21st (the longest), that triggered a trade closed on the fifth candle. I am right, or miss something?
22 February
trendy trendy
Please give me a hint...
22 February
fxtime fxtime
You are correct about the long candle on the 21st but look at what happened the next day...you got a mean reversion. In the blurb within the article it describes how long you can expect for these trades to complete and none in the attached chart failed you.
23 February
trendy trendy
"Three subsequent candles seems very high but the data reflects the maximum period and it is common to see a move complete in four hours."
Relevant articles for you

Disclaimer

The Saxo Bank Group entities each provide execution-only service and access to Tradingfloor.com permitting a person to view and/or use content available on or via the website is not intended to and does not change or expand on this. Such access and use are at all times subject to (i) The Terms of Use; (ii) Full Disclaimer; (iii) The Risk Warning; (iv) the Rules of Engagement and (v) Notices applying to Tradingfloor.com and/or its content in addition (where relevant) to the terms governing the use of hyperlinks on the website of a member of the Saxo Bank Group by which access to Tradingfloor.com is gained. Such content is therefore provided as no more than information. In particular no advice is intended to be provided or to be relied on as provided nor endorsed by any Saxo Bank Group entity; nor is it to be construed as solicitation or an incentive provided to subscribe for or sell or purchase any financial instrument. All trading or investments you make must be pursuant to your own unprompted and informed self-directed decision. As such no Saxo Bank Group entity will have or be liable for any losses that you may sustain as a result of any investment decision made in reliance on information which is available on Tradingfloor.com or as a result of the use of the Tradingfloor.com. Orders given and trades effected are deemed intended to be given or effected for the account of the customer with the Saxo Bank Group entity operating in the jurisdiction in which the customer resides and/or with whom the customer opened and maintains his/her trading account. When trading through Tradingfloor.com your contracting Saxo Bank Group entity will be the counterparty to any trading entered into by you. Tradingfloor.com does not contain (and should not be construed as containing) financial, investment, tax or trading advice or advice of any sort offered, recommended or endorsed by Saxo Bank Group and should not be construed as a record of ourtrading prices, or as an offer, incentive or solicitation for the subscription, sale or purchase in any financial instrument. To the extent that any content is construed as investment research, you must note and accept that the content was not intended to and has not been prepared in accordance with legal requirements designed to promote the independence of investment research and as such, would be considered as a marketing communication under relevant laws. Please read our disclaimers:
- Notification on Non-Independent Invetment Research
- Full disclaimer
- 沪ICP备13028953号-1

Check your inbox for a mail from us to fully activate your profile. No mail? Have us re-send your verification mail