CT Journal

[ChrisThong]

Have my first missing fill today and it is MIDD. My order was $173.27 and the low of the day was $172.28. My order was not filled. I called IB and was told that my order was part of the exempt trades, hence it was not filled. I have been assured by IB that it was nothing wrong with my account nor my order.

[JulianCohen]

What is exempt trades?

[ChrisThong]

The explanation I got from IB is this:

“Two counterparties to buy the same symbol and it was decided that my order was part of the exempted trades”

My interpretation of that is that they belonged to off-market trades.

[JulianCohen]

I’d understand if that was the low of the day, and the other guy got the trade so you didn’t, but your bid was higher than the low, so whoever sold at the low got ripped off.

You’ll never get a straight answer out of them anyway….just put it down to the cost of trading

[ChrisThong]

Exactly Julian. I checked 1min chart for MIDD and found that it opened at 173 and plunged to 172.28. My logic is that my limit price will get hit first before it gets to 172.28.

I have tried my best to understand what IB have said. At this stage, I would treat it as cost of trading and move on.

[JulianCohen]

Thinking about it, the trade was probably done in one of their “pools”, that’s why you didn’t get set but it has to be reported.

Shit Happens

[ChrisThong]

August 2021

US Market

MR#1 +1.75%

The above strategy started on 9 August 2021.

Also, today my ASX100 rotational strategy gone live, and I am trading this strategy with my superfund.

[ChrisThong]

A message for my future self.

Firstly, MR#1 and ASX100 rotational are in drawdown. Treat it like part of doing business in the stocks market and move on! As Nick says “Next 1000 trades”.

Secondly, I am in the progress of developing a second MR strategy and my first ever MOC.

In respect of the second MR, I hit a hurdle in the last few days and this is due to the chances (1610 or 16% for period 1/1/2005 to 30/7/2021) of the second MR system buying the same symbol as MR#1 (which is trading R1000). For this, I have tested a large number of ranking methods (thanks to the contribution from Said and Glen for sharing the codes in the forum). Until now, nothing is fruitful. From hereon, I will strip bare the second MR and test it again in the following fashion:

1. Adding price filter or any index filter to assess the win%;
2. Adding ranking method to assess the win%, CAR/MDD and Payoff;
3. Adding entry rules to have higher CAR/MDD and Payoff then metric achieved at item 2, above.

In respect of MOC, my first test achieved approx 32% CAR and 15% MDD, and I was like “woohoo I can trade this” then I realised that the account margin was set to 25, instead of 100 (no margin). Once, I removed margin from it the metrics dropped to mediocre level.

[GlenPeake]

For you 2nd MR system, I’d suggest also looking outside of the S&P500 and include taking a look at the R2000/R3000 and/or also the entire universe of stocks. (Price filters and Turnover filters enabled, would be what I would suggest if doing so).

I’ve been getting some nice outlier trades outside of the R1000 universe

e.g. 1 recent trade
BBIG

[ChrisThong]

Following on the discussion in Glen’s journal, and also, not to hijack Glen’s Journal, I have a detailed look at the trades of both MR#1 (trading in R1000) and MR#2 (Planning to trade in S&P500), both using the same ranking but different entry rules (after testing different ranking methods, thank you Nick for giving me some further ranking methods to try, I find using ROC for ranking gives better results and metrics so far). Test period is 1/1/2005 to 30/7/2021 for both strategies.

My findings are as follows:

1. Of total trades both strategies entered in the period, there were 1610 trades (16%) where the strategies would buy the same stock;
2. 41% of 1610 were losing trades with losses totalling $354,184.89;
3. 59% of 1610 were winning trades profit totalling $534,706.11;
4. The net effect is $180,521.22.

How I get the above numbers, I hear you ask?

Well, I run backtest for both strategies separately, and copy the trades into Excel. From there, I look for trades entered by both strategies (trades that have the same symbol and entry date). For the purposes of calculating the profit, I just summed up all the trades with profit more than zero. For losses, i summed up the trades with profit less than zero.

I hope that I am doing it in the right way. Please, let me have your thoughts on this if you do come across my journal. Thank you all and have a great day.

[JulianCohen]

Sounds right to me Chris.

I just did the same thing. I had similar results. Also I calculated the maximum risk of each trade.

Out of 18,000 trades I had 2 with a -4.5% loss and 8 with a loss of between -2% and -3.5%.

Not too bad really

[TrentRothall]

Hi Chris,

Did you run the test using compounding or fixed position size? eg 10k otherwise the numbers will be skewed tp the trades at the end of the test.

[JulianCohen]

I used fixed position size BTW…

[ChrisThong]

Hi Trent,

I ran the backtests with fixed $$ position size, so no compounding included.

[TrentRothall]

All good then, just thought i’d point that out

[Author]

Posts