Also, note that there are certain periods when the Rydex funds do outperform the QQQ, but overall that is not the case. I hope this is helpful for all. The question of the performance of the Rydex funds was in my mind for along time so I ran this study last weekend - surprised me!
Your observations about certain periods means that where you pick your dates to make the trades has an effect on the outcome, and that effect can be enormous. Prepare for a few more surprises!!!.
The RYDEX site has this Java chart you can set to different start dates and end dates to make comparisons to the NDX, or other indexes and funds.
rydexfunds.com 
If you open the link, you can then create a new chart using the set-up at the left to show simultaneously the RYVYX (Velocity), RYVNX (Venture), and the NDX (and more if you like). The slider bar can be adjusted in position and width. Grabbing the ends of the slide allows you to compress or stretch the window. If you set it to 2 days, you can position the start date anywhere you like, then grab the right side and stretch as much as you want.
As one example, if you set the slider bar to range from 7 October 2002 to 4 November 2003 you see the RYDEX fund has outperformed the Nasdaq 100 by much more than 2:1 (179.24%:77.64%; You can read the numbers by pointing to the curves).
There are time periods where the RYDEX fund seems to have done poorly in the longer term. For example in the interval from the December 5, 2001 to the high a few days ago, the NDX was down about 16% while the RYDEX was down 49%. This tends to supports your thesis. But notice that while the NDX fell about 53% to the October low, the RYDEX fell only 81%, and since the bottom turn the RYDEX has gained more than twice the NDX on a percentage basis. The problem is the RYDEX lost so much going down that the higher percentage gains on the way up are less in absolute dollars than the NDX. But don't forget that if you had a double long NDX and held it all the way down you actually lost more dollars in the decline. The reason it has come back farther in dollars is because you had more money at risk to begin with, and still had more money at work at the bottom, even though you had lost more in the decline. Had the decline continued, you would have an even greater loss compared to the RYDEX.
This "underperformance" of the RYDEX is not a matter of the fund absorbing premium; it is the nature of compounding the returns. The fund seeks to move 2 times the NDX on a daily basis. Over shorter time frames, the 2:1 ratio holds quite well. Over longer time frames, for a move in 1 direction the RYDEX funds do better than a double NDX. Whether the RYDEX does better or worse than the NDX is a matter of how you time your trades. Holding the 2x RYDEX through a long term reversal can be costly as the last example showed.
I took a careful look at your numbers, posted here again with fixed space font. I also added a line showing the compounded return instead of adding the percentages. The compounded return is the thing you want to compare, and for your table it makes things look even better for your argument.
Rydex QQQ QQQ x2
04/03/2003 Buy 65.32% 31.53% 63.06
03/31/2003 Sell -8.81% -2.15% -4.30
03/14/2003 Buy -2.80% 1.80% 3.60
01/17/2003 Sell -6.77% 1.08% 2.16
10/18/2002 Buy 7.98% 9.98% 19.96
05/28/2002 Sell 29.35% 25.17% 50.34
05/15/2002 Buy -10.11% -3.47% -6.94
01/14/2002 Sell 26.97% 19.49% 38.98
10/04/2001 Buy 53.39% 30.61% 61.22
06/18/2001 Sell 50.11% 27.30% 54.60
Added 204.64% 141.34%
Compounded 401.45% 249.63%
Your Rydex numbers agree with the numbers I get from the RYDEX chart, except for the first (bottom) one. RYDEX shows a return of 48.6% from 6/18/2001 to 10/4/2001. Your direct QQQ returns are impressive, apparently reflecting considerable trading skill. But there is, IMHO, a difference between trading skill and an inherent advantage in one of the instruments being traded. Apparently, you traded at various times during the day to get the returns in your table, but used end of day data for the RYDEX returns. I pulled the data for the QQQ from Yahoo. It is posted below. The November 14 entry appears to be what you used to calculate the last entry in your table.
Date Open High Low Close
14-Nov-03 35.81 36.02 34.95 35.04
3-Apr-03 26.64 26.85 26.31 26.48
31-Mar-03 25.60 25.73 25.21 25.25
14-Mar-03 25.73 25.98 25.36 25.72
17-Jan-03 25.90 25.94 25.27 25.31
18-Oct-02 23.28 23.85 22.93 23.81
28-May-02 31.47 31.52 30.49 31.10
15-May-02 32.08 33.63 31.85 32.81
14-Jan-02 40.49 40.68 39.70 39.96
4-Oct-01 31.65 33.20 30.94 31.30
18-Jun-01 42.64 43.17 41.40 42.03
I used these data to calculate the range of possible returns making the trades on the days you specified. As you can see in the table below, the difference between the best possible performance and the worst possible performance is very wide, and your performance is well above mid-range and well above what would have been achieved with end of day prices.
Date Worst QQQ Best QQQ Close QQQ
4/03/2003 Buy 30.17% 36.91% 32.33%
03/31/2003 Sell -5.67% -2.25% -4.87%
03/14/2003 Buy -2.96% 1.45% -1.83%
01/17/2003 Sell -2.81% 2.24% -1.62%
10/18/2002 Buy 5.95% 13.13% 6.30%
05/28/2002 Sell 21.78% 27.25% 23.44%
05/15/2002 Buy -9.33% -1.04% -5.21%
01/14/2002 Sell 15.29% 21.71% 17.89%
10/04/2001 Buy 19.58% 31.48% 27.67%
06/18/2001 Sell 23.55% 28.33% 25.53%
95.55% 159.21% 119.63%
To my thinking, a fair comparison of the QQQ to the RYDEX should be a comparison of apples to apples, leaving out your ability to capture above average prices by timing your intraday trades. That is the comparison one can make using the RYDEX data and charts. To that end, I extracted the data from the charts for the days you made your trades to build the following table:
Date Rydex NDX Close NDX x2
04/03/2003 Buy 65.32% 32.28% 64.56
03/31/2003 Sell -8.81% -4.46% -8.92
03/14/2003 Buy -2.80% -1.14% -2.28
01/17/2003 Sell -6.77% -1.26% -2.52
10/18/2002 Buy 7.98% 6.43% 12.86
05/28/2002 Sell 29.35% 23.19% 46.38
05/15/2002 Buy -10.11% -5.05% -10.10
01/14/2002 Sell 26.97% 18.25% 36.50
10/04/2001 Buy 53.39% 27.22% 54.44
06/18/2001 Sell 48.60% 24.35% 48.70
Added 203.12% 119.81%
Compounded 396.40% 187.30%
Note that the NDX closing data is very close to the QQQ closing data, and is more or less mid range of the best and worst possible returns, so I'm going to just use NDX closing data from now on. Note that on a compounded basis, the RYDEX funds return slightly better than double the NDX, but not by much. Note also that a few of the trades in your list heavily favor the NDX. In particular, the trade closed on 10/18/2002 for 29.35% was made after a huge drawdown on the RYDEX fund that had been up over 100% in early August. A skilled trader using the RYDEX funds might well have taken some profits earlier. I believe this table represents a fair comparison of typical trades made on these dates.
With perfect hindsight, I can tilt the tables in the other direction to illustrate that in a market with big percentage moves like those in the time period covered by your trades, the RYDEX funds have far greater potential than a double NDX. Changing the dates on some of the trades yields the following table (I even left in a couple of losers just for fun):
Date Rydex NDX Close NDX x2
04/03/2003 Buy 65.32% 32.28% 64.56
03/31/2003 Sell -8.81% -4.46% -8.92
03/11/2003 Buy 11.79% 6.24% 12.48
01/14/2003 Sell 25.43% 12.43% 24.86
10/07/2002 Buy 75.42% 36.07% 72.14
06/11/2002 Sell 49.84% 26.92% 46.38
05/07/2002 Buy -12.23% -5.03% -10.06
12/05/2001 Sell 93.77% 32.64% 65.38
09/21/2001 Buy 120.92% 52.71% 105.42
05/22/2001 Sell 176.90% 44.84% 89.68
Added 598.35% 234.64%
Compounded 5680% 626%
I'm not suggesting that I or anyone else could achieve these returns. That is not the point. The point is that the performance over time and several trades is enormously sensitive to the timing of those trades. One set of data representing 10 trades over a period of time that includes better than average prices from intraday timing of one instrument is not a valid comparison of the potential of two instruments being traded, nor is the last table that looks back for the best possible time to close some of the trades to favor the RYDEX. |
