Number of Holdings & Leverage
How the Number of Funds Held Affects the Final Outcome
-Using ETFs From the Mostly USA Portfolio-
with Market Timing
Percent Total Returns--Weekly--Hypothetical
versus Benchmarks
7.7 years: April 12, 2002--December 18, 2009
See Comments below
THREE'S BEST
The 'Rules' say :: Buy the two top-ranked funds. (And while we're at it, why not just use number 1, the top-ranked fund alone? If it's top-ranked, it should have the top-ranking performance, no? Logic says, Yes.) But the correct answer is counter-intuitive. It's, Buy the top 3 ranked funds.
For example, in the table below, if you hold 3 funds (top n column) when the market is in an uptrend, your total return (% tot return) is 2652% which is 53% per year compounded, giving to best gains of any number of funds held. The maximum draw down is -6.5%, which is 40% less than that of holding the top one or two funds alone.
EFFECT OF NUMBER OF HOLDINGS
ON RETURNS and RISK
Mostly USA ETFs Portfolio
Version 2.0
Hypothetical
Seven Years 2002-2009
Notes & Statistics below
|
%tot |
|
draw |
trds |
|
top n |
return |
%/yr |
down |
/yr |
|
1 |
1897 |
46.7 |
-10.9 |
9 |
|
2 |
2428 |
51.4 |
-10.4 |
12 |
|
3 |
2652 |
53.2 |
-6.5 |
17 |
|
4 |
2272 |
50.1 |
-6.0 |
20 |
|
5 |
2305 |
50.4 |
-3.8 |
21 |
|
6 |
2158 |
49.1 |
-3.6 |
24 |
|
7 |
2097 |
48.6 |
-3.0 |
27 |
|
8 |
2020 |
47.9 |
-2.0 |
28 |
|
9 |
1920 |
46.9 |
-1.5 |
27 |
|
10 |
1875 |
46.4 |
-2.4 |
27 |
|
11 |
1911 |
46.8 |
-2.2 |
25 |
|
12 |
1898 |
46.7 |
-2.1 |
24 |
|
13 |
1895 |
46.6 |
-2.4 |
22 |
|
14 |
1875 |
46.4 |
-2.9 |
20 |
|
15 |
1913 |
46.8 |
-2.0 |
15 |
|
16 |
1913 |
46.8 |
-0.8 |
11 |
|
17 |
1826 |
45.9 |
-0.2 |
7 |
|
|
|
|
|
|
SPY |
-1 |
-0.1 |
-49.1 |
-- |
top n is the number of highest ranked ETFs held in the portfolio. SPY reports the 'market' results. draw down is the worst-case draw down encountered during the entire seven-plus-years period that included the tail end of the 2002 bear market and the crash bear market of 2008. SPY is the ETF equivalent of the S&P500 Index. Its contrast with Mostly USA results seems an assault on credibility.
In fact, I find these results astonishing. The only alternatives are to report them as they stand, which I have done, or adjust them lower to levels more credible which I have not done and do not intend to do.
EFFECT OF LEVERAGE
ON RETURNS and RISK
Mostly USA ETFs Portfolio
Version 2.0
Hypothetical
Seven Years 2002-2009
Notes & Statistics below
|
%tot |
|
draw |
|
levg |
return |
%/yr |
down |
|
1 |
2652 |
53.2 |
-6.5 |
|
1.1 |
3799 |
60.5 |
-7.3 |
|
1.2 |
5392 |
68.0 |
-9.8 |
|
1.3 |
7404 |
75.1 |
-12.6 |
|
1.4 |
10354 |
82.9 |
-15.6 |
|
1.5 |
14946 |
91.8 |
-15.6 |
|
1.6 |
20345 |
99.7 |
-18.9 |
|
1.7 |
27593 |
107.8 |
-22.5 |
|
1.8 |
40019 |
118.1 |
-26.4 |
|
1.9 |
52589 |
126.0 |
-30.5 |
|
2 |
55254 |
127.4 |
-46.6 |
|
2.5 |
99402 |
145.5 |
-69.7 |
|
3.0 |
33793 |
113.3 |
-89.0 |
|
3.4 |
1975 |
47.4 |
-98.9 |
|
3.5 |
-390 |
ruin |
-100.4 |
levg is the leverage or multiple of margin used to increase the cost value of a position. Full margin in a brokerage account is typically 2x. Your basic cost of a position is 100% of purchase price. Your leverage position is two times cash cost.
Many ETFs today
offer two and three times leverage within a fund. These funds may be long or short (inverse). They are legal for IRAs and other tax-benefitted retirement funds.
The top red curve
in the chart above shows the best results when you buy the 3 top-ranked funds in an up market and sell short the SP500 or its equivalent when the market trend is down.
The current version of Mostly USA is considerably more volatile than Version 1.0. The volatility shows up in profits (much bigger than before). See Table of Results, previous version (July 2009). And in drawdowns (larger) and in increased number of trades.
Compare a previous chart of profit multiples with the current chart. The latter runs away on the upside. That's because of volatility and turnover. The turnover is due to a faster acting timing module in Version 2.0.
You can see that returns peak at 2.5x leverage, but the drawdowns have long since become intolerable. At 3x (which some ETFs make available), you are already on the road to ruin, which destroys you as you approach the 3.5x multiple level.
Yes, so where does that leave us? As to number of holdings, the top three ranked will make you more money, and for me is worth the extra work and time (we're talking here about a few more minutes a week instead of just a few minutes a week. Nevertheless, I will let the rule of 'top 2' stand. When the system was developed, I wanted it to be as simple as possible and require the least work to use it, consistent with above average performance and risk reduction. '2' did this better than '3' more easily and quickly. For myself, personally, I will likely use top 3. More important for the average user having accounts with smaller dollar assets, brokerage commissions could be too exorbitant proportionately to permit prudent use of the system.
Real-time returns records will continue to report on the basis of top-2 and bottom-1 funds for all purchases and short sales with timing.
Caveat. This study and these conclusions pertain only to Mostly USA ETFs. They do not necessarily apply to the Worldwide ETFs Portfolio. I do, however, plan to review and retest Worldwide early this year using the revised timing module present now in USA.
Draft posted
1/3/2010 6:08 p.m. ET
|
Back
| Subscribe
|
ET_STEP2/colsEI-ES
© 2008 The 2000 Corporation.
All Rights Reserved.
| |