Out of hours volatility?

[Just_me]

Hi there,

first post so please don't laugh too hard after reading my question:

I (presume) like everyone else trading End of Day, I've used historical data (Open, High, Low and Close) to backtest.

What this historical data does not tell me is how volatile the markets (e.g. FTSE) is out of hours (e.g. between London close and NY close).

So for example, you trade based on the closing price but in reality you're stopped out out of hours because of what's happening in the US.

There would be no way of knowing this from the historical data you have.

The fear here is that backtesting a "system" (I'm NOT selling or buying anything by the way) using this data would therefore be completely unrealistic.

Any ideas guys?

Thanks.

 

[tomorton]

Yes, this is a problem. Short-term if you're trading a market that is affected by another (e.g. the FTSE100, affected by the Dow and S&P), you can have some nasty surprises if the FTSE closed looking bullish, then the US went bearish after London closed. Of course, it can happen the other way too and you can have pleasant surprises when the US move ambiguous European markets unexpectedly in your direction.

You have a number of options though -
* trade the leading market (i.e. the US)
* widen or remove your stops overnight to allow for increased volatility
* backtest and monitor your system to ensure that your wins always outweigh your losses over time: don't forget that much textbook stuff on TA, price action etc. is based on US markets - its still of value but some generalisations don't necessarily translate directly to other markets
* modify your system to allow for joint TA of both the FTSE and the US - on the most basic level, ensuring that you are only long on the FTSE when NY is also bullish.

 

[Just_me]

Thanks for your quick reply Tomorton,

understand all the above, thanks.

One other question: what if you were to use the open, instead of the close?

In other words, enter at the open (as a proxy for yesterday's close) and exit at the close. Of course, this assumes no gaps between the open and the close, i.e. you can substitute today's open for yesterday's close (such gaps are much rarer for indices than for individual shares).

In the above scenario, the "risk" is being able to place your opening trade at the very first tick of the day which can be very volatile.

So my question is, in practice, how achievable would that be? In your experience can one get close enough to the opening tick, or are we saying that markets (indices) move so quickly at the open, that's it's pretty much impossible?

Thanks again.

 

[Xeno]

If you can get stopped out, you're still in the market and your backtesting should cover those times too. Instead of using eod close, use intra day prices.

If you can't handle overnight risk, you simply have to close your position.

 

[tomorton]

Hi Just_me - on the FTSE100 index, in practice there is no such thing as the opening price. FTSE themselves apparently take the previous sesison's Close as the next one's Open - which is of course nonsense, there clearly ARE gaps, but I think they do it because nobody can agree on what is the opening price anyway. Different sources give different Opens - some seem to take the Open as being the price after say 2 minutes of trading - but that is wrong as the market opens at 0800, not 0802. On a spreadbetting platform, the Open is readily visible, but it is the SB company's Open, not exactly that of the index itself.

In my backtesting I never rely on Opens for entries or exits, only Closes. In trading, I try to never enter or exit in the first half-hour of trading.

 

[tomorton]

PS - Many times, the overnight gap from Close to Open (or to Open plus about 30 minutes) is often the biggest move of the day, and often immediately starts to fill. If you're looking for this as your profit target, it might pay to get in late in the first session and exit with a very modest profit as early as possible in the following session. My rules above for backtesting and entries / exits are based on swing trading, which is what I do, not overnght gap plays.

 

[Just_me]

Thanks for your help everyone.

Look like the best options be to either take the overnight risk on the FTSE or trade the leading market (US).

Thanks again.