Friday, July 8, 2011

Market timing from a fellow blogger

Varan says at seekingAlpha:

If, at the beginning of each quarter, you had switched from SPY to TLT and vice versa depending upon which of the two yielded higher return during the prior quarter, you would have ZERO down years during the period 2003-2010, with only a dozen switches in the eight years, and a CAR of 14%.

Of course who knows if this will work in the future, but at least this year it has yielded 7.45%, pretty respectable.

------

Here are the results of this strategy for you to verify.
First column is the symbol (+ indicating a switch). The two dates are the first days of the week at the close of which the trading occurred, followed by the return. The last three numbers are the adjusted close data and the return for the prior thirteen week period on the basis of which the trade was made.

No need for crystal ball here but of course it is back testing.

You get still better results if you switch between VTI, EFA, and TLT.

All data based on Yahoo adjusted close data.

SPY+ 6/30/2003 TO 9/29/2003 5.11% 98.22 100.30 11.09%
SPY 9/29/2003 TO 12/29/2003 8.09% 100.30 110.10 1.62%
SPY 12/29/2003 TO 12/29/2003 0.00% 110.10 110.10 6.61%
SPY 1/ 5/2004 TO 4/ 5/2004 2.12% 111.69 114.46 6.87%
SPY 4/ 5/2004 TO 7/ 6/2004 -1.94% 114.46 112.37 2.36%
TLT+ 7/ 6/2004 TO 10/ 4/2004 5.48% 84.57 87.00 -0.82%
TLT 10/ 4/2004 TO 12/27/2004 1.90% 87.00 88.33 4.89%
SPY+ 1/ 3/2005 TO 4/ 4/2005 0.02% 121.56 117.36 8.26%
TLT+ 4/ 4/2005 TO 7/ 5/2005 7.52% 89.28 94.35 1.93%
TLT 7/ 5/2005 TO 10/ 3/2005 -1.86% 94.35 91.79 8.18%
SPY+ 10/ 3/2005 TO 12/27/2005 4.65% 122.96 126.96 1.84%
SPY 1/ 3/2006 TO 4/ 3/2006 1.26% 125.19 130.07 4.65%
SPY 4/ 3/2006 TO 7/ 3/2006 -1.82% 130.07 127.43 1.49%
TLT+ 7/ 3/2006 TO 10/ 2/2006 5.97% 83.90 89.10 -0.13%
TLT 10/ 2/2006 TO 12/26/2006 1.29% 89.10 89.45 6.93%
SPY+ 1/ 3/2007 TO 4/ 2/2007 3.04% 142.25 142.16 5.48%
SPY 4/ 2/2007 TO 7/ 2/2007 6.51% 142.16 150.87 1.44%
SPY 7/ 2/2007 TO 10/ 1/2007 2.36% 150.87 152.60 4.74%
TLT+ 10/ 1/2007 TO 12/31/2007 7.68% 88.69 92.78 7.13%
TLT 12/31/2007 TO 12/31/2007 0.00% 92.78 92.78 5.64%
TLT 1/ 7/2008 TO 4/ 7/2008 1.56% 94.18 94.62 8.48%
TLT 4/ 7/2008 TO 7/ 7/2008 -2.97% 94.62 92.07 1.06%
TLT 7/ 7/2008 TO 10/ 6/2008 4.48% 92.07 98.74 -3.06%
TLT 10/ 6/2008 TO 12/29/2008 22.57% 98.74 121.43 6.07%
TLT 1/ 5/2009 TO 4/ 6/2009 -8.44% 115.60 103.54 22.57%
SPY+ 4/ 6/2009 TO 7/ 6/2009 3.08% 83.34 88.94 -4.74%
SPY 7/ 6/2009 TO 10/ 5/2009 22.52% 88.94 102.90 5.24%
SPY 10/ 5/2009 TO 12/28/2009 4.46% 102.90 112.90 17.07%
SPY 1/ 4/2010 TO 4/ 5/2010 4.35% 112.37 118.25 4.46%
SPY 4/ 5/2010 TO 7/ 6/2010 -9.26% 118.25 103.64 2.83%
TLT+ 7/ 6/2010 TO 10/ 4/2010 6.09% 100.83 104.62 14.40%
SPY+ 10/ 4/2010 TO 12/27/2010 8.47% 114.37 125.13 6.73%
SPY 1/ 3/2011 TO 4/ 4/2011 4.95% 126.71 133.43 8.47%
SPY 4/ 4/2011 TO 7/ 5/2011 1.66% 133.43 133.78 5.18%
TLT+ 7/ 5/2011 TO 7/ 5/2011 0.00% 94.04 94.04 5.36%



(c) TonyP4 2011. Written in 7/8/11. Updated 7/8/11.

#####
Disclaimer: All my posts are for informational purposes only. I'm not a professional investment counselor. Seek one before you make any investment decision.

3 comments:

  1. Varan's strategy is fine and I'll find some time to test it out. It has entry and exit and I do not see any common pitfalls. We all use backtest data to verify strategies. Yes, there may be no holy trail in investing otherwise we are all sipping fancy tropical drinks in some beautiful tropical islands served by pretty tropical ladies...haha. It is far better to use strategy than without.

    I have a little suspicion as it is quite simple, but it does not mean it does not work without verifying it. One little book identifies the ROC works and it is even simpler. It may work, but after it is publicized it should not work. So I owe Varan for showing us his strategy.

    I've been creating and testing strategies for years for different market stages. It is hard to identify the market stage. It worked 2 years ago and I got 80% return for that year for my largest account. Hence I do not claim it will not work all the time. However, I will not show the details to the world unless I've to kill you.

    My strategies scan stocks for me and I do further analysis. I test whether my strategies/analysis are working currently. I keep it as a routine now. So far it works for me better than without strategies.

    My largest account beats over 90% of diversified fund managers in last 3 years. I do not claim it will continue to be so. I try to say that there are a lot of disadvantages of big fund managers and they become the market themselves. In last 10 years, the average return is about 2.5% (1% appreciation and 1.5% dividend). Taking 1.5% management fees would substantially limit their performance to beat the market.

    ReplyDelete
  2. Extremebanker says:

    Market timing as I use it has nothing to do about forecasting the direction of the market. It is about risk management. If you are a buy and hold investor then you put 100% of your investment at risk. Market timing with moving averages or some other clear technique (50 da crossing 200 da) limits the amount of risk you take because you will move out of the market when certain price levels are violated.

    A good portfolio manager enhances return and controls risk. To measure the enhanced returns with some market timing go to ETF replay and run a moving average backtest on SPY using a 200 day average. It outperforms and provides lower volatility. The return is over 5 times as great for the last ten years and the strategy drawdown was only 5.8%. Enhanced returns with reduced volatility.

    ReplyDelete
  3. Continued by Extremebanker:

    I started this process just after the 87 crash. I didn't have computers at that time so I used a percentage of the 52 week hi. I bought scudder growth and income. My definition of a bull market was if the market was within 90% of its 52 wk hi. If it fell below that I would go to cash. Usually a couple of trades per year.

    You may call this market forecasting. I was simply controlling the amount of exposure I was willing to take. A buy and hold investor places 100% of that investment at risk. I learned during the 87 crash I did not feel comfortable with that much risk.

    Well I did this for years and then I made an error and took money out and paid for two farms I purchased in 96. They have turned out good but not as good as the market would have done.

    Now we have computers and online trading so I use ETF's and passive indexes to manage macro investments. I also invest heavily in fixed income and manage an institutional bond portfolio professionally. I invest in DG stocks at work and personally but it is a subset of my total portfolio. I don't believe in trading. I try to buy for the long term but Mr. Market will not always let me do that. I think it is foolish to set and watch net worth evaporate in a savage bear market but you need a discipline before you invest the first dollar. Somthing simple and easy to execute. Having an investment plan to execute is the most crucial part of investing. It should detail how you will manage non systematic and systematic risk. Risk management is the crucial element to successful long term results IMHO. The simpler the plan the easier to execute. It doesn't have to be complicated. I believe investors should strive for absolute returns rather than relative returns although I have never owned enough short positions to be profitable during a raging bear market such as 2008.

    I think John Serrapere best describes my approach. Give me 75% of the up markets and less than 50% of the down markets. That works for me.

    ReplyDelete