Video

Peter Garnry
Dax pushes through 12,000 for the first time since April 2015. The price momentum is supported by a strong Eurocoin growth indicator, PMI figures and Ifo survey data, Saxo Bank’s head of equity strategy Peter Garnry explains.
Article / 13 July 2016 at 13:00 GMT

Successful trade opened? Probably, but go shopping first

Hypothesis Testing
United Kingdom
  • Trading has some similarities to retail shopping and queue hopping
  • We all know when the markets we trade are likely to be volatile, or slower
  • Time of day can influence whether a trade is profitable
  • Flow-on effects from Asia  can influence early trading in European markets
By fxtime*

There are a surprising number of methods for calculating probability. 

Some are are mere experience events. For instance, you look at a queue of shoppers at various tills and determine which queue is probably the quickest to clear. We can from experience determine a multitude of factors that would influence our choice – which shopper has the largest purchase going through or ages of shopper that may slow down the transaction. Who has distracting influences to slow the process such as very young children, or even the presence of a trainee cashier at the register.

So experience teaches us a version of quantification which we have all been wrong with. We failed to see a flaw in our queue selection. Once we join a queue and wait patiently we might have an additional influence that occurs that would have changed our original selection. A change of cashier occurs whilst you are queuing or an earlier customer finds an error with their purchase. All of the above are real time events that we respond to and transact with.

nnn
 Various factors can influence the time we spend in a queue 
and the queue choice decisions we make. Photo: iStock

How we handle the changes involves yet more choice. Probability would suggest you stay with your original selection because the cost in time to move all your purchases to another queue and re-assess how fast each queue is moving and so on is detrimental to your objective of moving through the process quickly.

Trading is to some no different to humble shopping. We open a trade and await its outcome. This is trading probability through experience. Certainly there many who can (time and again) be consistent enough to succeed and grow their trade account. But the vast majority will not.

So how do we determine a numerical value to our probability of success? Will it enhance our ''odds'' of winning?

Let's reconsider the shoppers in the above example. Would our transaction be simpler by shopping at a quieter time of day? Would we find shopping for our weekly fresh groceries simpler on a Thursday as opposed to a Friday or Saturday? Changing the simplest part of the transaction can enhance the speed of the whole process.

Trading is no different. We all know when the markets we trade are likely to be volatile, or slower. Depending on our trade structure, for example, options require volatility or time decay to really earn for us, then we select our timings in accordance with this.

Typically time decay trades perform better 53 down to 45 days away from expiry for maximum potential profit, but volatility is often an opposing contra trade when volatility is at its busiest such as selling the VIX when the market is at max panic or if you like when all the tills are working and have maximum queues !

But that is just experience again. Sure have some quantification but that is for option players not CFD. Let's puts some numbers to our craft.

I mentioned that we use experience in the above commentary, but let's ask how much experience.
  • Determine exactly how many times you have done what ever event.
  • Furthermore exactly how many times was the trade successful?
From these two simple pieces of information, we have the ability to say as a percentage just what success rate we have for opening and closing a trade in profit.

Let's be honest a percentage of having a successful trade isn't enough for any trade as it doesn't tell us the real history.
  • How many trades were successful is relevant.
  • But what profits were made?
  • How much of a drawdown occurred?
  • What margin did you use in relation to the profit gained?
  • How many consecutive trades (maximum and minimum) were successful?
Then look at the opposite...
  • What losses were incurred?
  • How frequent were those losses...did they occur consecutively?
Again we can improve our percentile overview. Could we reduce the stop requirement to ensure our profit margin increases or would this mean a greater frequency of losses?

Now we need to consider the timings of the trade and consider whether there is any correlation to your trade timings to the trade results. Are more trades successful on certain days with less drawdown like say the shoppers above seeking their weekly groceries on a Thursday as opposed to a weekend?



nnn
 Source: Wikipedia

See this earlier article as well: Adding bell curves and whistles to your trading strategy.

Morning skew

Bayes Theorem on probability is based on the likelihood (or if you like the probability) of an event occurring given that another event has already happened. So we look at our data and see we have made a series of trades in the mornings and see the markets moves beyond say the second standard deviation level more frequently than the afternoons.

Because there are greater influences to the market, we have the follow on effect from Asia moving the European marketplace and the opening market moves of Europe first thing as they readjust to world pricing events. In other words we have the equivalent of weekend shoppers altogether pushing market prices up or down. Which using Bayes theorem suggest we have the highest probability of success. (due to market skew!).

Frequency is the fastest determination of probability. It is low level mathematics that is usually regarded as the real time acid test as to whether something works.

nnn
Follow-on effect from the Asian trading day can influence European markets at the start of trading, as markets adjust to world pricing events. Photo: iStock

S&P500 performance


We know that 53.1% of the time, the S&P500 will rally or move sideways (using 5 years/10years/15 years and 30 years of data this figure/percentage has remained robust) and frequency also tells us that the minimum move was 2.5 points. So a simple long at the close would 53.1% of the time be successful.

However as a trade structure this is useless as drawdowns would be corrosive. And losing days fall far further than the potential gain. The average drop is 11 points. So now we have a problem. We can lose on average more than we can gain....out of every 100 trades we have;

  • 53.1 succeeding with an average 2.5 points profit, which equals 132.75 points
  • 46.9 trades failed with an average 11pts loss = 515.90 points lost

Frequency obviously highlights the problem.

However frequency also gives us the answer for resolving the losses. We will know from our frequency data that losing trades occurs a maximum of two consecutive occurrences on extremely rare occurrences three times. So if we waited for two consecutive duff (that is poor quality) trades, and then started trading, we can expect a profit.

But how much profit?

Frequency tells us that the average run of successful trades is seven consecutive days, so we trade for that with expectation that on average we will gain 7 x 2.5 points equals 17.5 points profit, then we await the two duff trading events before resuming our trades. On average we do not incur losses. However they still will occur as no trade model is perfect so how do we minimise the effects further?

Frequency will show that the S&P500 will drop often by a maximum 8 points before it recovers to make 2.5 points profit for us, due to overnight market moves. Thus place your stop at 9ts so we ensure our stops are tighter than the average losing day drop.

Again probability and frequency can improve your profits. As on balance more days are bullish than bearish we can stagger our trades. Open half stake size as normal seeking the 2.5 points and then open another at 5 points lower, both with exact same stop level as described. The 5 points lower has the highest frequency for dips so is the optimum level to enter and enhances profits up to a sum total of 5 points

  • For example, 2.5pts x half stake plus 7.5pts x the other half stake size = total 5pts.

Finally probability shows that a move beyond 9ts dip is bearish and the market has further to fall which gives us a whole new array of strategies but that's a whole new series of frequency data to re-assess.

Next time you visit your supermarket consider probability. If you are a scalper, you want a wift in and out transaction and will often seek the mispriced market value. So you will seek out the end of line bargain(s) a position player will buy on a Monday or Thursday for efficiency. Relative value traders seek economy of scale; eg wine bought by the box not just a random bottle purchase. Look and you will see all sorts of real time shopper/traders LOL.


*fxtime is an Alias

— Edited by Robert Ryan
13 July
fxtime fxtime
Whilst we see the classic bell curve chart above normally I turn it through 90 degrees. Then assume the close price is the central value line of the orange classic curve...equally I assume the central 50 line of the RSI 5 matches the same line as the close price. Thus at the end of the day you can overlay the true RSI position to have an early indication of market bias....the closer to the central line the more likely we will gain a change in trend. We already know the standard deviation levels and high probability trades from them so we need to apply the short term skew using the close value of twoprior days prior compared to the current close price to assess where an impulse move of high probability can be positioned.
13 July
fxtime fxtime
Remember probability can be determined visually as a probability cone and this too can be applied to show direction...profit levels and trail stop margins.
13 July
zefy zefy
My trade sheets confirm morning skew and higher probability of success in the morning.
13 July
zefy zefy
It became obvious when I started better documenting details of each trade.
13 July
zefy zefy
Excellent educative article again, and entertaining too!
13 July
fxtime fxtime
LOL...thx Zefy. Next may need you to expand on the programming side :-)
13 July
fxtime fxtime
Following on about probability perhaps this is something to consider;
https://www.tradingfloor.com/posts/drowning-in-ice-cream-correlations-6867790
I will enter a short at the close ! Remember this is a probability play we could get another rally day but then I would enter towards the close as per this trade plan!
13 July
zefy zefy
Yes! I have 5 raising day lows strategy finally implemented to my system too and was noticing 4th day this morning.
13 July
fxtime fxtime
The last few of these have been fun :-) Brexit was great but I traded very small stakes as when I saw the day count I was very concerned about the trade set up but stats models ignore World events so you have to follow the trade plan !
13 July
fxtime fxtime
Zefy you must have quite a programme data trading base by now :-)
I am impressed.
13 July
zefy zefy
It is single code base which is run for all symbols.

int raising_lows = 0;
for ( int i = 0; i < 100; i++)
{
if ( Low.LookBack(i) > Low.LookBack(i+1) )
raising_lows++;
else
break;
}

"Platform stuff" and trading rules are separated to ease introduction and testing of new rules.

Overlays are sorted already by the C# code so I avoid using spreadsheets:

_coretable.DefaultView.Sort = "Value desc";
_coretable = _coretable.DefaultView.ToTable();

C# also produces standard format csv for input to R so I do not need to change R script between trading models.

I have learned huge amount of new ideas and rules from your articles.
13 July
fxtime fxtime
Love it...neat and efficient stuff...smart stuff and especially like the separation of platform & trading issues. Always best to keep well defined.
14 July
fxtime fxtime
I shorted at the close but was stopped out by the break of yesterdays day high which is annoying. Now the market has rallied even further to the current 2172.00 in globex cash market price and I have re-opened a new short now !
14 July
fxtime fxtime
Well the SP rose to the three day 3rd std dev but has retraced nicely so far and I am covering after the BoE inspired drop for +10 at 2162.00 and will await the USA opening before any further SP trades.
14 July
fxtime fxtime
Correction ...I have just checked my account and I covered at 2163 so only made +9 ....oh well.
14 July
fxtime fxtime
If I can get 2172 - 2173 again on SP500 then I will re-enter a short at 50% stake size and look for an opportunity to scale the balance later....we may well be seeing a capitulation move today.
14 July
zefy zefy
Thanks for updates. No trades here yet for this as I have had very strange problems with my account today.
14 July
fxtime fxtime
Was it due to the downtime in data feed for SP500? The Saxoplatform and TorontoDominion Bank sites were down at the exact same time.
14 July
zefy zefy
Don't know, probably not... FGBLU order was cancelled without any reason and E7U was filled at ridiculous price levels... After that I cancelled everything because could not understand what was going on...
14 July
zefy zefy
SP is holding still... For me tomorrow must be better day :)
14 July
zefy zefy
Too much energy spent today for account issues...
14 July
fxtime fxtime
Yeah when things go pear shape it is best to step back and await the next day. Hopefully everything is sorted for you?
FWIW as we haven't dropped below yesterdays day low then I will short again near the close ! We could rally to Monday morning as we have expiry to get rid off and that can stall matters whilst they are all closed/covered before any meaningful move occurs.
15 July
zefy zefy
Yes, moving on now. I got somewhat distracted yesterday LOL. But lesson learned, I guess it is obvious for everyone, that it would be safer to trade forex than currency future. Nevertheless, I am watching todays expiry and see if there starts any move after.
15 July
fxtime fxtime
LOL...the algotrade setup article I have forwarded for next week; makes comments on that very topic !
15 July
fxtime fxtime
zefy are you watching the eurgbp algo? The ratios have altered after the 10am time stamp point....eurousd ratio to gbpusd is the stronger therefore eurgbp should rally which is what we see now....so far its in profit :-)
15 July
fxtime fxtime
FWIW as the ratio triggered on a 2:1 ratio but only just....so the opening trade is smaller and will be scaled as the ratio improves. Stops are as previously commented.
15 July
zefy zefy
Yes, I am watching it. Actually completed the implementation of ratios today morning. Now considering practice trade with very small stake.
15 July
fxtime fxtime
Ratios are almost neutral now so caution as it seems the CPI today will have a big bearing on this. My stops will not permit negative ratios and so far I am only +5 in profit so a close call so far.
15 July
fxtime fxtime
stopped out for +4 so will reset at the CPI release.
15 July
fxtime fxtime
Did you put the six lines ''in'' for the audjpy as opposed to just the two line ?
15 July
zefy zefy
All six lines, like in your example.
15 July
zefy zefy
Next time will think twice before taking shortcuts LOL.
15 July
fxtime fxtime
Great stuff....may be I should post my algo programme here to compare but I will await outcome of next weeks article first. FWIW my system permits a reset of time fix period so I can reset the algo around news events like the CPI release later today but these are just tweaks to make life simpler.
My old boss used to always say there was a flaw in everything there is always a mote that needs to be quantified. I always worked on the two line entries and realised that I was getting odd trade results at times and then developed to the four and still found quirks but find the six line scenario removes most of these.
Whilst next weeks article is an intro to algo assesment and one programme I will probably post one a week after that for intraday/daily/weekly and swing....lets see what how readers take it.
15 July
fxtime fxtime
CPI data is out and ratios have turned to the eurgbp long
15 July
fxtime fxtime
ratios about to be neutralised again so I am covering now at +1 only. Looks like one of those days of whipsaws grrrrrr as it is late in the week now I have de-activated the algo as don't want a good week spoiled by a duff trade.
15 July
zefy zefy
Yeah, I was watching that too.
15 July
zefy zefy
SP was not impacted by CPI...
15 July
zefy zefy
At least immediately :)
15 July
zefy zefy
I was hoping it turn to red but it is so persistent :D
15 July
fxtime fxtime
It will imho. I have a short still running so am happy enough to leave it for the moment.
15 July
zefy zefy
Me too!
15 July
fxtime fxtime
Zefy I have just got back to my desk and see the suggested eurgbp was a good call and I didn't do the trade aaaargh I hope you got the +65 available LOL?
15 July
zefy zefy
Yeah, but very small stake. Practice trade and not all 65..
15 July
zefy zefy
Good call from you anyway.
15 July
zefy zefy
Meanwhile still watching SP.
15 July
zefy zefy
Starts to show interesting development, not sure relevant enough though.
15 July
fxtime fxtime
Wow well done...for the week the algo should have gained +425 but because I de-activated the function earlier I only made +365.
Interestingly the AUDJPY function earned more :-)
Probability curve suggests Monday for a drop....anything today is welcome though :-)
15 July
zefy zefy
At least color is now right!
15 July
zefy zefy
Sure :)
15 July
fxtime fxtime
It is.....can I ask have you tried colour coding your trade strategies/algo's? Not sure if you noticed but when I post my charts sometimes the horizontal lines are different colours? Where-ever I position the line for a manual trade scenario or order the selection of line colour tells my systems what sort of trade to open and likewise what sort of stop is required !! Fun times :-)
15 July
zefy zefy
Open positions have green/red background for long/short bars. System generates text string/table with price levels and action for those :).
But I do not trade from chart because I am better reading text instructions than charts LOL. If I need to change something, I go investigate code and after that possibly chart.
15 July
zefy zefy
Meanwhile SP is developing :)
15 July
zefy zefy
There is a lot to improve in my system to automate it more. It is nowhere near complete!
15 July
zefy zefy
I mean platform. Trade rules are of course forever ongoing development project :).
15 July
fxtime fxtime
Sounds like you are very advanced already mate......bleedin' brilliant even :-)
As for ongoing development...that will never change we all have to keep adapting and enhancing :-)
15 July
zefy zefy
Thanks. I keep eyes open for new ideas and TF site has proven to be excellent place for new ideas and views and comments.
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

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