Saturday, December 31, 2011

Final 2011 Stats

These are not my actual trading returns. Mine are listed in a chart below. The following stats are compiled using raw signals from my system, using TNA as a trading vehicle. Results below do not account for slippage, taxes, or commissions.

Annual Compound Return: 51.43%

Total Trades: 46
Winners: 35 of 46 (76%)
Avg Return per Trade: +1.12%
Profit Factor: 1.63

Breakdown

Long Trades: 27
Winners: 20 of 27 (74%)
Compound Return (Longs only): + 27.76%
Avg. Return: +1.03%
Avg. Return (winners only): + 5.01%
Avg. Return (losers only): -8.28%
Worst Loss: -44.36% (ouch)
Best Gain: +11.90%
Profit Factor: 1.32

Short Trades: 19
Winners: 15 of 19 (79%)
Compound Return (Shorts only): + 35.41%
Avg. Return: +1.86%
Avg. Return (winners only): + 3.55%
Avg. Return (losers only): -4.45%
Worst Loss: -8.36%
Best Gain: +8.69%
Profit Factor: 2.90

All returns above incorporate whether the system was 50% or 100% invested. For example, if the system went 50% long at TNA 40 and sold at TNA 44, it does not get credit for the full 10% gain.

At first glance, it appears my system did better shorting the market and I suppose that is true. But if you play the "what if" game and pretend that 1 bad trade in August didn't happen, my system would've returned well over 110% for the year, the majority of which came from the long trades. Remove that 1 trade and the profit factor for all long trades would've jumped up to 6.60!!! (Which is an insane number). I introduced stop losses in August, so eventually this weekend I will calculate some stats assuming I had stop loss rules in place all year (done - below).

Here's a fun chart tracking my actual trading account throughout 2011. You can see I couldn't keep up with the 51% performance of my system. But after August, I was happy just to finish in the black. I remember praying several times back in Aug. Looks like it worked! :)











Hypothetical August Stop Loss case. CAUTION...almost unbelievable results ahead. :)
If we had the stop loss rules in place in the summer, half of the long trade initiated in July would've been stopped out on July 29th when my model flipped to unhealthy status. TNA closed at 72.71 that particular day. The other half of the trade would've been sold at 44.15 on Aug 11. Stopping half the position at ~72 results in moderate damage versus severe damage like we experienced. The trades that followed in mid-late August were nothing short of amazing (as can be seen in my chart above) and the total loss in August would've been a paltry -6.61% versus the actual -29.44%. Obviously this makes a world of difference. The account at the end of August would still have been up over 41% for the year versus the actual 6%. The difference by the end of the year would've put the system up 100.4% through December versus the actual 51.43%! Like I said, almost unbelievable. So far, my system has flipped to unhealthy status before every recent crash. The process by which a market tops before crashing leads me to believe this stop loss feature will protect assets in a similar fashion in the future. The outlook appears bright.

All-in on the first signal scenario...WARNING...UNBELIEVABLE RESULTS AHEAD :)
Taking a 50% position on the first signal, followed by going all-in on a second signal was a terrible strategy in 2011. While playing with smaller amounts of capital might have lowered my stress levels, it substantially lowered profits... SUBSTANTIALLY!!!
Here is a profit curve with a $10,000 initial investment.










Your eyes are not deceiving you...that is a 268% gain for the year!!!

Late November had back to back losing trades that were pretty big, but by far the biggest blow to the system came in August when it gave back almost 50% on that 1 trade. Can you imagine if I plotted this curve using the stop loss rules? Let's do it!









The loss in Aug is marginalized and the system returns 387%!

I'm going to give my 50% / 100% strategy a few more months to prove its worth before I completely scrap it. But with a profit curve like that, I can't see sticking with it much longer. Especially now that stop losses are built in.

Market Health Section
I've described how I determine market health several times. I only use market health as a qualifier for my signals, but that doesn't mean it can not be used for long term buy and sell signals. Here is a chart showing when my model flipped from healthy to unhealthy and vice versa. Assuming a long only SPX strategy with no leverage, it would be about a 5% gain for 2011.


Last but certainly not least - Strong buy and sell signals.
(Defined as when both healthy and unhealthy modes are generating the same signal.)

There were 29 signals that were strong (either buy or short). There were 23 winners, 6 losers for an 79% win ratio. Of the 6 losers, 5 were very mild. The 1 major loss came in the summer. Check out this profit curve assuming you only traded strong buy/short signals using a 100% positioning strategy (TNA).













That's a 200% return despite losing almost half the equity from 1 trade in July. Looking closer at that bad trade, the strong buy signal bought TNA at ~80 on July 14th. A week later TNA was up around 85. So the buy signal was decent, the problem was my system failed to generate a sell when it should have. Neither buys nor shorts had an edge. All strong signals (long or short) should be followed. One possible strategy since strong signals are rock solid is to use 100% positioning for strong signals and 50% for regular signals. Food for thought...

Thanks for reading!!! Let's see if the system can keep producing these kinds of results in 2012. Insert standard disclaimer here about how past results can not predict future performance, and trade at your own risk, etc, etc.... Happy New Year and God Bless!!

To view all the raw signals, click below. On the ensuing page, scroll halfway down and click the gray "Download This File" button.
2011 Signals Excel Spreadsheet
Sorry these links are broken!!! Email me for files!

As a fun extra bonus, here are the old 2010 signals. Keep in mind, the system was still in its infancy in 2010.
2010 Signals Excel Spreadsheet
Sorry these links are broken!!! - jtrader21 at gmail dot com

26 comments:

  1. Thank you for this wonderful website.

    One question: is your return based on a 100% position with each trade, even if, for that particular trade, only 50% was actually invested? In other words, is the return for the total amount that is available for investment or for the actual amount invested with each trade?

    ReplyDelete
  2. As you were typing that question, I was typing the answer above... :)

    When I get finished with all the stats, I'm going to investigate how the results would've varied with an all or nothing strategy.

    An all-in or all-out strategy sure would make calculating these stats much simpler!

    ReplyDelete
  3. The stats are very interesting.

    It would indeed be nice to see how the system did with an all-in or all-out strategy.

    Take care and thanks.

    ReplyDelete
  4. Ok, that's enough for today. Tomorrow, I will post the fun stats including the all-in/all-out results and the "what if I used a stop loss in August" results.

    Both will be interesting.

    If there are any other stats you'd like to see, actual or hypothetical, let me know and I'll try to accommodate.

    ReplyDelete
  5. what is considered to be a full or 50% position. Is the full position 100% of the money available to invest or a smaller % of the total I have to invest?

    ReplyDelete
  6. 50% or 100%, those are impressive numbers indeed. You beat both my real returns and my "fantasy" ES trading returns. Good job! I think in 2012 I'll start trading your system :-)

    And Happy New Year!

    ReplyDelete
  7. xtrends,

    100% is whatever you are willing to risk. 50% is just half that amount. I start each position as a half position just in case I don't hit the top or bottom of a move, leaving plenty in reserve to average in at a better price if the opportunity presents itself.

    If you want to follow my system, take a comfortable amount in your account to allocate toward my signals, whether it's 1,000 or 100,000 doesn't matter. When I get a signal, use half of that amount, and if I get another signal, use the rest.

    Rolling the gains from one trade into the next will compound your returns and help your money grow faster.

    ReplyDelete
  8. Thanks, Michele! Happy New Year to you also.

    I don't want to mislead anyone here into thinking I actually made 51% gains in my real trading account. Sure, I could have if I followed my system blindly to the letter. I didn't quite have that kind of discipline in 2011. My resolution for 2012 is to do a better job of that.

    The problem with that is it requires you to shut everything and everyone off completely. Reading and researching will only serve to bias one's trading, something that will severely hamper your willingness to follow a system blindly.

    When you love trading and enjoy researching, it's difficult to turn the computer on at 3:45pm, get the signal, place the trade, then turn the computer off for another 24 hrs. But that's what I believe you have to do to follow my system or anyone else's.

    That is going to be a challenge for me.

    ReplyDelete
  9. What were stats for strong buys and sells? Also what would results be if only used the first 50 percent buy and not averaged in on the second?

    ReplyDelete
  10. I think that JTrader is onto something very important. There are all kinds of system out there, some, like his, excellent, some good, some okay, etc. But for a system to work, a trader has to turn off that little voice saying not to take the current call because of a million different reasons. What happens is that every once in a while the little voice is right and then one keeps listening and going against the system. But, in the end, a good system will, in its methodical and emotionless way, outperform 99% of discretionary traders.

    So, even though it's hard, it's important to give a system a chance by following it, especially when the little voice is saying not to. If, after a while, the system's equity curve starts to look terrible, then you can move on. But you shouldn't mess with a system until you've given it a chance to do its thing unencumbered by second-guesses.

    ReplyDelete
  11. Thanks for the words of wisdom Algyros.

    straubw07, Thank you for reminding me about strong signals. I'll take a look, probably on Monday.
    For you second question, I just answered that in the post. See the section titled All-in on the first signal...

    ReplyDelete
  12. That's some pretty heady stuff, J-Trader. I'm glad I have an extra day this weekend to digest it all.

    Thank you very much for all your hard work.

    ReplyDelete
  13. J-trader, could you post the dates that your system switched from Healthy to Unhealthy and Unhealthy to Healthy, for all of 2011? Thanks.

    Chuck

    ReplyDelete
  14. J...thanks for the analysis. That is awesome stuff. Will follow your trades this month and let you know. Also, I had asked you about the Indian market. Can you share your spreadsheet so that I can try to incorporate Indian market numbers ( index and VIX ) and see how it works ? I will only paper trade and also back test for a few months. My personal email is padysrini@gmail.com.

    Thanks

    ReplyDelete
  15. J...why not start posting a SL column on the left side ?

    ReplyDelete
  16. J-trader
    Thanks for the very complete analysis and statistics.
    Your information and data are impressive and well support your trading methodology.
    Looking forward to strong results in 2012!

    ReplyDelete
  17. Jim, thanks, and my pleasure...I too am looking forward to trading 2012. If nothing else, these stats have hammered home the fact that I should not be deviating from my system... ever. :)

    golfdude,
    Stop loss section has been added. Good suggestion.

    I will post all my signals from 2011 and 2010 on one page so you can easily compare to the Indian market. You'll just have to remove TNA values and plug in the ETF/index of your choice. Thanks.

    ReplyDelete
  18. J=Trader,
    Do the strong signals get generated at the same time as the weak signals? In other words, do you know about strong signals earlier than weak signals, or do they materialize at the same time?

    ReplyDelete
  19. About the same time. Although it really just depends on the amount of intraday volatility. Big market swings in the last hour of trading can (and often do) throw me for a loop.

    ReplyDelete
  20. I was wondering if you had the ability to backtest your system further back than 2010? I know that TNA didn't exist ten years ago, but SPY did and one could extrapolate returns. It would be an interesting way to see how the system behaves in different kinds of markets.

    ReplyDelete
  21. I'd be especially interested to see how your system handled bear markets, like in 2008.

    ReplyDelete
  22. "When you love trading and enjoy researching, it's difficult to turn the computer on at 3:45pm, get the signal, place the trade, then turn the computer off for another 24 hrs."

    It's difficult for a person, but it's real easy for a computer. I'd like to write a program that takes your signals and places the trades automatically.

    Do you think you could come up with some simple standardized format for your calls (for example, "BUY TNA 30.50 LIMIT") that would be easy for a machine to parse? I think it would be fun to try this out.

    ReplyDelete
  23. One more thing: thanks for posting the .50 stop loss. However, it would be much more useful if you could translate that into TNA.

    ReplyDelete
  24. It goes back to 2008 but I used 2008-2010 data to build the system so returns those years benefit from data mining/curve fitting. In addition, other indicators I use go back only a few years as well. There are too many unknowns to go farther back, so my system will just have to keep proving itself in the present. So far, it's doing a fine job. I don't plan on going anywhere, so my hope is to build up years worth of results.

    Half stops on long trades simply use a recent swing low. I don't see why TNA price wouldn't work.
    Things are a bit more complicated if the system is short in an unhealthy market as the flip to healthy isn't as straight forward.

    Michele, do you mean just putting a line similar to that in each of my posts? Sorry, I'm still not sure how that stuff works...I'm trying to learn though.

    ReplyDelete
  25. Thanks for the information, J-Trader.

    Like you, I didn't follow the signal last week. Another lesson on not over-thinking systems. I'd be interested to know what you plan on doing (going long now or waiting for a pullback that might never come).

    ReplyDelete
  26. I'm sitting on the sidelines until the close at which point I will most likely be going short. I don't expect a huge sell off today after such a violent gap up, so a buy signal is probably off the table.

    ReplyDelete