[ScottMcNab]

I wouldn’t take a system live without at least a month of paper trading. The paper trading is a further test that there are no errors in either the system code, the exploration afl or the exploration for the api. I found that paper trading the system taught me almost as much about the system as designing it did.

[ScottMcNab]

Is this in relation to my first post Maurice ? I was referring to the stress testing….I was wanting to see that the metrics that I had for system were not degraded significantly by the different stress tests listed at the end of the course. For me I guess this meant I would still consider system acceptable with worst results returned from the stress tests.I guess another way of saying this is that the CAR must not drop significantly and the maxDD must not increase too much…so in essence, yes !

[ScottMcNab]

36 buy limit orders placed with BatchTrader for paper trading acct for MRV

[ScottMcNab]

Avira had an issue with TWS a few weeks back…TWS suddenly shut down on opening every launch when had worked fine for many weeks….seems fine again now

[ScottMcNab]

Piggybacked some of Nick’s orders over the last few days and found my slippage on (sell) mkt orders lot higher than his…turns out this is essentially a reflection of inherent limitation of a paper trading verses live account…so will move on from slippage issue which has consumed my attention for a while now….. but will have to monitor closing when start live trading

Live trading delayed while seek further advise re setting up necessary structures needed for RegT margin acct

Picked up Nick’s API for buy limit orders to manage MRV system….so next step is to learn this new piece of kit

[ScottMcNab]

I found with the MRV it was interesting to do backtesting/stress testing by changing

LEPrice = Min(Open,Ref(buylimit,-1));

to

LEPrice = Ref(buylimit,-1);

This should be the worst possible fill with limit order….was an interesting test of the system….helps give some confidence in the system without proving too much I guess

[ScottMcNab]

I still struggle understanding some of the code that people post so don’t sweat it…the saving grace is the looping templates Craig and Nick have created and posted towards the end of the course work….using these (and substituting in different criteria for buy and sell without changing the template) has enabled me to develop systems that seem worthwhile to trade…I’m sure you will find it the same. I had no real coding experience at the start of the course.

[ScottMcNab]

The stats you have posted on your MRV system previously were impressive Said. I have not seen/ don’t recall the stats for the other systems. A question I have been thinking a bit about recently is whether the advantage of adding multiple systems to diversify (minimize maxDD ?) harm long term CAR. I guess the simple solution is to have two systems of reasonably similar CAR. I haven’t seen too many trend following systems approach the results of your MRV system though. Was it a choice to accept a reduced portfolio CAR to minimize portfolio maxDD that led you to using multiple systems Said or are your other systems similar to the MRV CAR ? (not after specific results of other systems….interested more in your thoughts on the topic to help clarify the issue in my own mind).
Thanks in advance for your time and thoughts
Scott

[ScottMcNab]

Any way to tweak it so enter on open next morning ? May lose some profit but gain from being able to use ranking system may make it worthwhile ?

[ScottMcNab]

Thanks Nick

[ScottMcNab]

Went over trades for the week

28 buy at mkt (at open)
31 sell LOC

average cost of commission per trade on buy: $1.78
average cost of slippage per trade on buy: $25.88

interestingly the average slippage on the sell orders was in my favor….so on very limited data may balance out ok….will continue to investigate reducing slippage on buy orders

[ScottMcNab]

This would indicate that IB not so much wanting ppl to use directed API orders ? Cant imaging it costs them more….may be it reflects commissions IB receives by routing orders in certain ways…beyond me …but added this commission into the backtester and it knocked 3% off CAR…this may more realistically reflect future results though than having uncertainty in open price

[ScottMcNab]

Still digging to learn more about the variation in the fills I’m getting on open with mkt orders compared with Amibroker open prices…OHLC prices from NYSE and NASDAQ websites match amibroker data for relevant stocks so the issue may lie with me selecting SMART as routing for the mkt order… I may in fact be getting opening price but it is simply not the NYSE opening price (which is the data Amibroker uses for NYSE stock)…so live/future trading may more accurately reflect my backtesting if I select NYSE in the order ticket rather than leave it as SMART….will test

[ScottMcNab]

Working late Thurs night…no luck with replacement yet

[ScottMcNab]

From what I can gather its the market on open order for IB…..OPG being the abbreviation for opening…..I just placed tomorrows buy orders manually as I’m too ignorant to be able to change dde api….so I better go read up on which column to change on dde api . I will try and post the results for the 7 buy market on open orders tomorrow…my fills against “official” opening price again so can compare with mkt order…based on my track record there is a good chance I’m barking up the wrong tree though

[Author]

Posts