Mad Money Machine

by Paul Douglas Boyer

Some Lazy Portfolio Entries Backtested

From the 66 entries in the Lazy Portfolio Smackdown game, I have taken these 19 portfolios in which all component funds existed back on 1 August 2000 through 31 January 2008 and computed their Return vs. Risk profile. To see the components of each portfolio, go to the Portfolios page and find the ID number for the portfolio. [Note, in some cases I substituted regular index funds for Admiral shares, mutual funds for their ETF equivalent, or other similar but not identical funds… such as VGSIX for FIREX in one case. And on lew’s portfolio I cheated and started PCRIX (PIMCO CommodityRealRet Strat Instl) on 1 Jan 2003.]

ID Lazy Portfolio Return Risk
1 CyberBob 4.39% 6.80%
5 Jarrod 7.81% 11.30%
8 james22 6.26% 14.11%
10 mikenz 7.90% 9.47%
13 poulinbob 16.68% 19.57%
16 HAZEL 7.64% 4.89%
17 Justin617 7.98% 10.64%
19 sgr000 7.12% 10.30%
22 Sunny 4.24% 7.78%
24 MrBG 7.15% 8.36%
33 DR 4.52% 14.10%
38 lazyrad 5.05% 15.37%
39 gflippin 3.70% 5.38%
41 Kevin Ucker 12.67% 13.89%
45 Bev and Mike Cote 7.56% 4.99%
51 Jennifer & Mark 11.40% 10.62%
54 lew 10.69% 13.20%
59 Cosmo 5.02% 2.35%
60 TnGuy 11.10% 13.37%

The graph below plots the values from the above table. The best place to be is top and left. Note how you could almost draw a line from bottom left to top right. That would be the efficient frontier line. The idea is that the more risk you take, the more you should be rewarded. If a winner were to be awarded in each of three risk bands (0-8%, 8-16%, and 16%+) based upon this graph it looks like HAZEL, Jennifer & Mark, and poulinbob would each get a copy of Index Funds: The 12-Step Program for Active Investors. Although Kevin Ucker and Bev and Mike Cote have it going on too.

image

It is worth comparing these results with the Professionals I showed in a previous posting.

And just to see what you would have gone through in the past one year, here is how the portfolios did since the start of 2007:

ID Lazy Portfolio Return Risk
51 Jennifer & Mark 23.86% 9.89%
60 TnGuy 13.15% 13.66%
13 poulinbob 12.10% 16.37%
59 Cosmo 8.18% 2.05%
45 Bev and Mike Cote 7.86% 3.51%
16 HAZEL 6.21% 3.40%
54 lew 5.88% 11.56%
22 Sunny 3.61% 6.77%
1 CyberBob 2.98% 7.09%
39 gflippin 2.67% 5.29%
10 mikenz 1.21% 9.43%
38 lazyrad 0.57% 13.08%
5 Jarrod 0.10% 10.55%
33 DR 0.03% 12.91%
24 MrBG -0.05% 7.33%
17 Justin617 -1.44% 9.74%
19 sgr000 -1.48% 8.92%
8 james22 -2.18% 11.78%
41 Kevin Ucker -14.23% 12.12%

Fri, February 8 2008 » Blog

2 Responses

  1. wizkid February 9 2008 @ 4:26 pm

    OK – Ready to audit the calculations.
    Please define the calculation of Return and Risk using the multiple periods.
    Is return – average annualized return, compounded return with reinvesting dividends, rebalanced annualized return, etc. etc.
    Is the risk based on monthly returns, monthly rolling etc.

    Explain the reason for the timeframe – why Aug 2000 as the start date?

    Great job on comparing the portfolios!

  2. Paul Douglas Boyer February 9 2008 @ 9:32 pm

    Oh excellent! Love a 2nd pair of eyes.

    I started with August 2000 because if I went back further in time, VIPSX (Vanguard Inflation Protected Securities) didn’t exist prior to that (at least according to Yahoo). And it is used in several portfolios.

    I pull monthly “Adjusted” historical stock quotes from Yahoo finance (using that free tool RCHGetYahooHistory) These presumably account for reinvested dividends and for things like splits.

    I calculate but do not report the annualized return and annualized risk for each fund. Then I also calculate and report the total portfolio’s return and risk, as below:

    Five columns are used in the spreadsheet for each fund:
    Date, Adj Close, Normalized, Gain, Value
    Date = the first trading day of each month
    Adj Close = the funds value brought back from Yahoo including (subtracting) reinvested dividends
    Normalized = first month, fund value starts at 1. Each subsequent month therefore shows the total gain
    Gain = delta in the value between each month. This is used to calculate the annualized standard deviation.
    Value = this fund’s contribution to the total portfolio. Basically, multiply the Normalized column by the fund’s percentage in the portfolio

    Then to compute the returns and risk on the total portfolio, I add up all the Value columns for each month, calculate another monthly gain difference column for the total portfolio, compute the total annualized standard deviation, and report the annualized return as:

    =+(Last value/First value)^(1/((End date-Start date)/365))-1

    I calculate Annualized Standard Deviation as
    =STDEV(RANGE*SQRT(12)) where RANGE is the gain column (each monthly difference in value).

    It takes about a minute for the spreadsheet to load as it draws all the stock histories from Yahoo.

    I DO NOT PERFORM REBALANCING. I should probably do this, huh? Will have to think about how to code it up in Excel.

    Suggestions for improvement greatly welcomed.