actually regarding the topic of missing entries i modified my MOC recently so the entry condition is not Low <= Limit price but L +0.01 < Limit price
this will eliminate from my backtest all the trades that i bought on the low of the day or 1 cent away from the low

this way my backtest is more pessimist but i know what to expect, regarding slippage i believe over time they balance out sometimes it is with you and sometimes it is against. Honestly i do not check them

I get them and it is frustrating when they occur. I have found that once in a while I do get a fill on these, but pretty much always a partial fill, so there is a small glimmer of hope there. I count it as part of doing business.

They happen about once maybe twice a month for me so I didn’t bother to record them. To put that into perspective I did 55 trades on my MOC last month, and that was a slow month. I had one of those issues. I’m trading the Russell 1000, and I chose not to trade the 2000 nor the 3000 as the number of missed trades would exponentially increase.

I see trading a larger universe and having a larger number of positions as a good way to combat this by helping to not have a system that depends on profits from ‘long tail’ trades (especially if these large quantity orders to be filled in a short amount of time at the open are the long tails).

You want your system to be making profits in a large number of trades right in the meat of the distribution curve, not have it only become profitable if the long tail gains/losses are included. I see the long tail gains/losses as an artifact I need to accept in ensuring the meaty part of the curve is where I make the money. I would much rather a system that performs 10,000 trades in ten years across smaller positions and makes the core profits from a boring, low return on 9,800 of those trades in the meaty part of the distribution and put up with 200 trades in long tail wins/losses along the way that don’t end up skewing the results much, as compared to say, a 1,500 trades in 10 years system that might be making a little more profit every trade or in fact depending on the profits of the 200 long tail trades to be viable.

Losing/skipping/missing trades in a high volume environment where each and every trade becoming absolutely, critically, accurately, correctly filled (or not) becomes more of a “whatever” moment when they do get missed. High volume and a 50 dollar problem as opposed to low volume and a 2000 dollar problem so to speak. This also helps if you were out too late tonight having dinner and drank too much and forgot to put your trades on or you were in a plane and had no access for 24 hours.

Julian, understood that you don’t wish to trade a broader universe as the likelihood of these slippage/missed fills increasing, however I think it does not exponentially increase as a problem at the same rate the universe broadens (or at least my experience hasn’t seen an explosion in these errors with a broader universe). If anything, a broader universe and more positions helps combat this problem to the point where if you do miss a trade, it is much less relevant and you care less.

For example, perhaps a 20 position system (forgetting leverage for a minute), in addition to the amount of funds you are using (let’s say 200,000 USD) and also what you have set for minimum share price (Let’s say 10 USD), this would see each position at 10K USD and on a 10 USD minimum that’s 1,000 units to buy at open.

Now assume 50 positions and everything else the same (10 USD min, 200K funds) this would see you needing to buy each position at 4K USD and therefore just 400 units to buy at most at open.

Obviously needing a quick fill on 400 positions is going to have a much better chance of happening than needing a 1000 unit fill.

Earlier in the year I was facing the same problem. I was telling myself (and trading) systems with a smaller number of positions (e.g. 15 positions) and thought this was preferable to systems with 50 positions. Along with the leverage my systems were hoping to try and get fills on 5,000 to 20,000 units at the open on one symbol. This was a real nightmare because as we know you get the fill in buckets. You might get a 600 unit fill, then another 380 units, then another 1600 units then 850 units (etc etc with seemingly random bucket size until the full 15,000 unit order is filled at which point 60 seconds has passed and the price is no where near the open and your average fill price does not reflect your target).

Since then I have learned my lesson and now trade two systems with 60 positions each and half the funds each and I’m now needing a fill on about 3,000 units at the absolute most on the very lowest price symbols, and usually it is more like a few hundred units on average that I need to fill so usually I’ll get an instant fill in one bucket. On the occasions where I now get a partial fill (or a missed fill altogether) I am not too worried about it as it is only a small miss in a much larger pool of tens of thousands of trades over time. In the hundreds of trades I’ve done since redesigning the system, I’ve probably missed altogether or had partial fill five or six times. And they are only a few hundred dollars missed here and there, and some helped me and some to my detriment, as compared to missing perhaps thousands of dollars missed here and there before, and on a much more frequent basis (before, in a month of 300 orders and 30 fills I would miss a fill or get a partial fill perhaps 5 to 10 times and the damage was horrible. Now, in a month of 2000 orders and 250 fills I would get a partial fill or a missed fill altogether maybe five or six times and the damage is negligible).

Something else I have done to reduce slippage and give myself the best possible chance of getting rapid fills is by running TWS and API on my server in New York which performs my trades (I trade only US markets, no AU markets). I don’t use a VPS, I use a bare metal dedicated server because I believe this is better for a few different reasons (which I’ll share if you are interested).

Hey by the way, I’m sorry but I realise now I probably shouldn’t write the way I do here in the forum. Me saying things like “You want your system to be….” and “This is better than that….” is probably sending the wrong message like I’m an expert, know it all, telling you what to do etc.

I’m not trying to preach or state any of my ways are better, just trying to give an angle from my perspective. I guess my brain dump onto paper doesn’t work well sometimes.

I’ll be more careful to express my opinions as opinions only next time…

Cheers,
Matt

i also trade R1000 for MOC and slippage is rare thing and can be ignored.
regarding trading R2000 or R3000 i agree it provides larger universe and higher number of fills but there is one main issue for me, during backtest i assume i am hitting 4 times leverage 40*10 while in reality not all stocks can be leveraged 4 times some are only 2x some 1x some 1.5x
since these data are not found in the backtesting stage this is why i avoid these two groups.