Quote from zouy2000: Hi Rol, Your system looks to me like some kind of mean reversal system and the result is simply great. However, I have two concerns, First, you seem to use 2x overnight buying power, can you handle some kind of black swan event which may be much worse than 9-11 and the market simply gapped down 30%. Do you have enough time to adjust your position to meet the margin? Hi zouy2000, A black swan event is obviously a concern, if not the biggest concern of a RTM strategy. Warren Buffett has a famous quote that goes: âBuy when there is blood in the streets.â Baron Rothschildâs was also quoted, âBuy when thereâs blood in the streets, even if the blood is your own.â They made a fortune having the guts to invest when others were afraid to. If you are looking for a contrarian opportunity, research shows that the most rewarding investments are made when the entire market is in the tank as it was right after the crash in 1987 and right after 9/11. The Whoâs Who list of millionaire/billionaires that have brought home plenty of bacon thanks to their investing acumen when most investors were cowering include the familiar, such as Warren Buffett, Benjamin Graham, Peter Lynch, Sir John Templeton and Martin Zweig. While my style is not "investing", I believe it is just a matter of time horizon. My period is days, while with an investor it may be years. Nonetheless, the market patterns are quite similar between the two. I have a spreadsheet in Excel to perform various "worse case" scenarios to see the effects on buying power during extreme market drawdowns. I place limits on the maximum number of positions I can hold, so I won't get margin calls. Ideally, you want to have maximum market exposure at the bottom. A gap down is actually superior to a slow bleed down, where you may be trying to scale-in slowly. I would prefer that the markets tank 30%. Individual stocks are having flash crashes and black swan events all the time. You cannot predict when they will occur, but you can predict what you will do after they have occurred. Second, Is your system scalable? Suppose you have one million equity, do you have enough signals which you can enter? You may have some liquidity issues at some point. Here are the number of buy signals I received recently just following slightly fewer than 1000 names: 3/8 18 3/9 28 3/10 143 3/11 58 That is 247 entry signals in 4 days. I could easily track closer to 2000 names, which would have put that number at around 500 entries. So for example, if I bought 1000 shares each of a $30 average stock price: 1000 x 30 x 500 = $15,000,000. The trick is not to be in the market all the time, but wait for panic to set it. I liken it to doing a "drive by" on the market; get in and then get the heck out. Sometimes, I imagine myself as a well-prepared fire fighter. They run into a burning building when the public is running out.
I played around with Excel charts this weekend and came up with this plot of my annual return compared to the SP500. I will post these on a monthly basis, to track my performance. It is M2M, being based on my real-time account net worth. 2/18 was the recent market top, and I took some heat for being a bit over-leveraged. However, it just pulled back to being a bit below break even on the SP. On the close 3/3 and open 3/4 my system had moved mostly to cash, enabling me to take advantage of the 3/10 and 3/11 lows, by buying dips in some oil related names, metals, and semis. This is why there is a divergence at the close on 3/11. I have currently positioned myself with moderate market exposure to take advantage regardless of which way it heads, meaning if it goes up I sell, and if it goes down I buy.
Just out of curiosity, I entered data into my chart going back to 11-15-2010. For the most part, it mirrors the SP, but showing more volatility. 2X leverage and being in individual equities likely accounts for this explosive nature.
I do not recall my system placing so many buys and sells concurrently, but that is what was happening today. A position would stop out for a small gain, opening a slot for a new entry in another position. The exits today netted me $181.68 on the day and my account net worth is at a new high for the year. However, given that the Dow futs are down 250 right now, I do not believe that will hold. I have decided to loosen my entry criteria for the time being within the strat since it seems the markets have changed since the mid February highes. Sell 3/14/2011 14:59 Buy 3/14/2011 11:39 Sell 3/14/2011 10:36 Buy 3/14/2011 10:11 Buy 3/14/2011 10:10 Sell 3/14/2011 10:08 Buy 3/14/2011 9:58 Sell 3/14/2011 9:35 Buy 3/14/2011 9:34 Buy 3/14/2011 9:33 Buy 3/14/2011 9:32 Sell 3/14/2011 9:32 Sell 3/14/2011 9:18 Sell 3/14/2011 9:06 Sell 3/14/2011 8:43 Buy 3/14/2011 8:41 Buy 3/14/2011 8:34 Buy 3/14/2011 8:34 Buy 3/14/2011 8:32 Buy 3/14/2011 8:32 Sell 3/14/2011 8:32 Buy 3/14/2011 8:32 Buy 3/14/2011 8:32 Sell 3/14/2011 8:31 Buy 3/14/2011 8:31 Buy 3/14/2011 8:31 Buy 3/14/2011 8:31 Buy 3/14/2011 8:31
How has your pure paper trading compared to your live? In your live account you have to adjust to match your bp as well as ignore some signals because you cant take them all. But the paper trading doesnt have to do this. Also your returns is using 2x, so would it be more accurate if you doubled your spx returns for an accurate comparison. Hence YTD you have marginally outperformed spx. But your paper trade results from may to may last year is impressive. You should track your live vs paper, and see where and why it diverges. Because it doesnt seem there is huge alpha considering the 2x overnight.
How has your pure paper trading compared to your live? In your live account you have to adjust to match your bp as well as ignore some signals because you canât take them all. But the paper trading doesnât have to do this. Hi DGunz, Many thanks for the thought provoking questions you provide! I havenât compared the back testing for a couple weeks now. I will do that when I get home from work tonight. Amibroker actually does ignore signals when it has reached its BP limit. You can code margin BP limits into the portfolio strategy. That is actually a big advantage that Amibroker has, to make it more realistic. That being said, I will never be able to match it exactly because my real life code in TS does not match the Amibroker code exactly. When my TS strategy gets stopped out to prevent a decent gain from turning into a loss, Amibroker does not do this. One reason for this is that Amibroker is not looking at intraday data, and another is that I havenât figured out how to code it. It has been more of an observation I have made over historical data that it should help in preventing bigger draw downs, although it may hurt my overall longer term performance. I would prefer a smoother equity curve, even at the sacrifice of some hypothetical long term profits. As I have said before, there is no shortage of signals, so it may actually improve long term results. I have given up on trying to match Amibroker results exactly. I am content that if I am doing well when it is doing well, and doing poorly when it is doing poorly, then that is enough for me. It actually takes more risk than I am comfortable with right now, because I have it maxing out margin buying power. My strat is not the product of curve fitting over a short time frame. I have tested in 100âs of time frames and it has always held up relatively well. The strat is based on genuine behavior I witness over and over again in the markets. Also your return is using 2x, so would it be more accurate if you doubled your spx returns for an accurate comparison? Hence YTD you have marginally outperformed spx. I donât use margin all the time. In fact I would say I spend less time using margin than I do using it. Maybe a 1.5X or less comparison might be fair. I mainly want to see if my strat is able to avoid big drawdowns which would cause me to underperform the spx . I view margin usage as my âtrump cardâ during trading. You only pull it out when you need it. But your paper trade results from may to may last year is impressive. You should track your live vs paper, and see where and why it diverges. Because it doesnât seem there is huge alpha considering the 2x overnight. I need to study up tonight on what alpha, beta, theta, etc. means I have seen these terms used a lot in these forums. I am happy to say that my Account Net Worth is only down $173 from yesterdayâs high for the year, even after todayâs nearly 300 pt drop in the Dow, if that is what low alpha means.
You are saying that your long only strat basically didnt flinch at all m2m from yesterday to today? $173 is nothing the market is down about 1% and your strat only lost $173 m2m? That is incredible, if I think I know what you are saying. Alpha just means that you are out performing vs a benchmark. If the market is up 4% from when you started and you typically use 1.5 leverage, then technically buying spy with 1.5 leverage would have yielded you 6% vs your performance of 9%. Which I am not knocking, thats perfectly fine.
Here is an open long position showing a stop placed at around 11:15am by the strat that nearly got hit today.