Pondering my future after working on an intraday trading strategy lately, I realized that I have absolutely zero idea about how to handle scaling. The majority of the systems I have designed before were multiday/week/month positions, so scaling was not as much of a factor -- I could build positions over time. Slippage didn't factor in. Time was on my side. However, when designing an intra-day equity or futures system, I have run into a wall, wondering if scaling truly becomes a major factor -- all of a sudden, I wonder if I can capitalize on the amount of capital I have available. So my question to all of you is this: how do you handle scaling? Is it a factor you take into consideration on system design? Do you run multiple, non-correlated strategies? Do you work in multiple markets at once? How is it possible to take advantage of large amounts of capital with small time-frame trades? Also, at what point does 'scaling' generally become an issue? I understand that this question is loaded, given that it depends on the strategy being traded -- but I would love some insight. Let's say I was doing simple linear scaling -- buying an extra contract for every X amount of dollars extra I had. At what point does the cost of this scaling outweigh my position benefit? Is it as simple as estimating cost versus benefit and finding the maximum contracts the system is capable of trading (ceteris paribus, of course -- let us assume our risk management strategy minimizes draw downs as a constant factor of capital)? I realize that it isn't really applicable to most small-time traders (so please, no 'if you have to ask, you don't need to know' type answers) -- but I am wondering how large funds tend to deal with this issue. Thanks so much!
Scaling in or out are strategy components that should be discarded on every time frame. Thank you for your time.
You completely misunderstood my question. I don't care about scaling in and out of positions -- I care about scaling strategies to take advantage of capital increases. Not all strategies can be run equally with $100,000 as they can be with $10,000,000.
Renaissance Technologies seems to trade futures and stocks, intraday, with tens of billions of dollars under management. And it appears that they are relatively successful. You could consider that an existence proof: yes it is theoretically possible, yes it is actually occurring right now, today, in the real world. Be of good cheer. It may be important to choose your trading instruments with large-scale trading in mind. It's probably much easier for a 200 million dollar account to dip into and out of the Bund futures on EUREX than the Orange Juice futures on NYBOT. I imagine that Renaissance trades lots more contracts of Bund than contracts of Orange Juice. You might want to explore the idea of trading many different instruments at once; spreading out your risk among (let's say) twenty different products, and reducing your betsize per trade by a factor of twenty. Allowing your account size to scale up by a factor of twenty.
MGJ, I appreciate you following me around these forums and answering all my questions. I knew the answer was out there -- I just didn't know in what form. I assumed diversifying into multiple markets and products would allow a firm to better utilize their capital ... but I was wondering if any other forms of effective scaling existed. Thanks!
In my case I trade my main intraday strategy only on the array of sp500 products. This system does work on several other products, but, I programmed/designed it to work specifically on SPY, ES and the associated options. Come to think of it, from the first day I started coding, I approached the issue backwards, i.e. my aim was to automate a system that trades profitably only in very liquid products.... I recommed you do the same. At this time, I am nowhere near moving the market, but, I do get slippage. I've heard - and if someone has any ideas on this let me know - that 1-2% of a products daily volume and you're pretty much not moving anything, some have said it's up to 10% for anything intraday. I have a hard time believing that 10% WON'T move a market significantly. Todays SPY volume was 400mil, 1% is 4mil - no way could I get 4mil shares in 5 mins without running the price... no f---in way. I'd be very interested in anyone who's done a study on this... Mike Edit: That being said, maybe .1% of daily volume is reasonable. 400k shares of SPY would probably be do-able, but one could easily spread it over the ES and the options. 400k * 128.50 = $51.4mil. As a rough guess one could probably day trade 50-100mil on the SP products without making much noise.
If you are referring to the 95% traders who lose money, then the question of scaling or not scaling is moot