Intraday Systems Flawed

Hi,

Id be very interested to hear of your experiences with intraday systems and their success.

I have been a professional systems trader for a couple of years, and have always traded medium term (holding winners for at least 3 days on average).


There is obviously a huge temptation to develop intraday systems, no overnight exposure, lower trading margins and the ability to compound returns will increase your account very aggressively.

However i dont see how you can realisticly overcome the slippage costs. I think most people are WAY too conservative on estimating slippage. A lot of people getting involved in systems trading dont even consider slippage!!

Obviously it varies from market to market, but take the bund for example, a liquid contract and with recent volatiltity has had some big ranges (in excess of 100 ticks). Now say your intraday system trades twice a day. You should allow for at least 2 ticks slippage a trade, this includes comissions and bad fills, which happens far more than you think, stop market orders often get swept and even worse if you use limit orders it can trade you price and not fill you before running off for a 15 tick winner!!

So assuming two trades a day 2 ticks slippage a trade thats 40 (in this case Euros) a trade your system will under perform your backtest. Thats about 800Eur a month per 1 lot you trade! (quite an expensive way of doing business im sure you'll agree) or Eur9600 per year! (remember this is for every 1 lot traded, costs go up exponentially)

I know you could argue "so what does that matter if you're making 30,000 per year per 1 lot" quite true. But for a system to have that much "edge" and have longevity i think is a big ask.

This isnt a "know-it-all" point, just my experience, can anyone prove me wrong?

totally agree while i would say you can of course make money intraday and some people do. I think your point is extremely valid as i think there are far too many traders who join what they think is the intraday gravy train without actually working out the slippage costs and another important factor is the spread they are paying maybe bunds is not too bad but in fx if your scalping for say 20 pips and your spread is 4 going in and 4 going out thats a huge % of what your looking for. This is a taboo subject brokers and large market particpants don't mention to newbie traders as they need them in the market to provide cannon fodder and easy liquidity. If they actually presented the real odds of success including the spread and slippage would anybody bother to trade off a 5 min chart? i wander? Just my opinion from 5 years of trading in futures and some fx spot. Certainly an interesting discussion point anyway!

thanks for the reply littleian30, hopefully more people will join in.

Your point about FX is particularly valid as this is where most retail traders start (losing) money as barriers to entry are very low. However you said (assuming 4 pip spread) cash FX you would pay 4 pips getting in and 4 out. This isnt so, if the spread is 50@54 you buy 54's then have to sell 50's to get out. So you just pay the spread once, not twice, but your point still stands.

Also agree money can be mde intraday. I started off as a market maker in the the energy markets and that really was a gravy train, never seen such easy money (while it lasted) the desk i worked on would make thousands of trades a day and rarely had a down day, but that was mainly because we could buy the bid sell the offer in wide illiquid markets i.e were using slippage in our favour

yes your right my mathes doesn't add up with the spread thanks for correcting it.

i think most traders don't ever do a basic simulation where they take their model backtest and add just 2 pips slippage and see the diference it makes to the overall p & l. If i had a £ for every good model i have had before i added spread and slipps i would be richer than Carlos Teves Agent.

yeah you and me both, so disheartening to build something that ticks all the boxes, smooth equity cuve, good profit to max draw, high percentage winners etc etc add 2-3 ticks slippage and it doesnt make a penny

the ability to compound returns will increase your account very aggressively.

Click to expand...

Or reduce it just as aggressively in my view.


Paul

tommog said:

. . . but take the bund for example, a liquid contract and with recent volatiltity has had some big ranges (in excess of 100 ticks). Now say your intraday system trades twice a day. You should allow for at least 2 ticks slippage a trade, this includes comissions and bad fills, . . .

Click to expand...

Unless you're doing something stupid like trading over figures you can shift at least a couple of hundred bunds trading "at market" (assuming you've got direct access) with no slippage in that contract.

FWIW, I've been making money trading fully automated intraday only systems since 2003. If you build the models right, they overcome execution costs. I'm not sure what else to say, other than to suggest that your conclusion might be based on reasoning of the "I couldn't do it, therefore it is impossible" variety and that with continued effort you might prove yourself wrong at some point.

To Magicmagician

Mathemagician said:

FWIW, I've been making money trading fully automated intraday only systems since 2003. If you build the models right, they overcome execution costs. I'm not sure what else to say, other than to suggest that your conclusion might be based on reasoning of the "I couldn't do it, therefore it is impossible" variety and that with continued effort you might prove yourself wrong at some point.

Click to expand...

i don't think we said it wasn't possible at all! to the contrary infact we said it was possible but we were making the point that its not as easy as newie traders beleive it to be and also that people should be more aware of these slippage costs and impacting trading performance.

you should re-read peoples posts carefully before barging in to proclaim how clever you are.

LOL! OK, whatever.

FWIW:

tommog said:

Hi,

Id be very interested to hear of your experiences with intraday systems and their success...

Click to expand...

Unless you're doing something stupid like trading over figures you can shift at least a couple of hundred bunds trading "at market" (assuming you've got direct access) with no slippage in that contract.

Click to expand...

Unfortunately i've had very different experiences, i have traded the bund systematically for several years and it is not uncommon to get a tick slippage getting in and out. If i have a stop order to buy 30 bunds if a 1 lot trades at 57's (for e.g) and 5 other people have a similar order you'll be lucky to get that price, not to mention the very likely scenario that somebody doesnt lift a 1 lot into your price but 50lots +.

Also, as Littleian said, the point of this thread was just to see how people have faired systems trading intraday considering your ratio of slippage to win size is undeniably higher. Its great to hear people are doing it

Littleian30 said:

i don't think we said it wasn't possible at all! to the contrary infact we said it was possible but we were making the point that its not as easy as newie traders beleive it to be and also that people should be more aware of these slippage costs and impacting trading performance.

you should re-read peoples posts carefully before barging in to proclaim how clever you are.

Click to expand...

Do I detect a Napoleon complex there "Little" Ian ?

pedro01 said:

Do I detect a Napoleon complex there "Little" Ian ?

Click to expand...

anything useful to add to the discussion about Intraday systems and slippage Pedro?

Any chance of an ashes lull in the market at 11 am i wander?

Littleian30 said:

anything useful to add to the discussion about Intraday systems and slippage Pedro?

Any chance of an ashes lull in the market at 11 am i wander?

Click to expand...

5' 3" would be my guess

pedro01 said:

5' 3" would be my guess

Click to expand...

i just can't see any point in either of your comments Pedro they don't relate to the thread at all? have you nothing better to do than scan through these forums and make silly remarks?

Have you made any good trades today to speak of?

Littleian30 said:

i just can't see any point in either of your comments Pedro they don't relate to the thread at all? have you nothing better to do than scan through these forums and make silly remarks?

Have you made any good trades today to speak of?

Click to expand...

Have you just started using the interweb my vertically challenged friend ?

Just curious, that's all...

Productive contribution to the thread ?

you should re-read peoples posts carefully before barging in to proclaim how clever you are.

Click to expand...

FYI - short ES 880.75 -> 878.75 for 2 contracts, stopped out at break even on the 3rd.

Not mechanical now - I have come to prefer the 80/20 rule to 'mechanical' trading - do 20% of the work to lead me to the trades & then make the decision on entry myself.

at least i was making an attempt to discuss what the thread was about! you have offered nothing at all about the subject of the thread.. which begs the question of what are you doing in it accept to write silly messages!!

Do you have an opinion on intraday slippage and spreads and how these can impact your own trading models? if so please enlighten us?

do you mean the internet? the interweb i have never heard of?

well at least you managed to get a trade on congrats.

May I suggest that you two trade insults via PM instead? Let's get back on topic, eh?

tommog said:

Unfortunately i've had very different experiences, i have traded the bund systematically for several years and it is not uncommon to get a tick slippage getting in and out . . .

Click to expand...

Going in you have a target price yes? In that case make it a limit order. If you get a fill you get a fill.

Going out then use stops.

Depends on how the sysem is constructed I guess, some systems (ie set of ruels) would allow you to pre-emtively place "limit if touched" type entry trades.

It also's gonna depend on how much you look to take on each trade, I'd say anything less than 5tick then the entry/exit points are important, but if you look to capture 5+ tick swings then take it on the chin.