No. of Recommendations: 5
"But I can prove that conventional passive-indexing strategies depend on assuming that returns distribute normally and that making that bet is irresponsible if 'responsible' means 'being able to quantify one's odds'."

Technically, the models work for any stable distribution, not necessarily only a normal distribution.

I maintain that indexing is valid as long as one understands the risks and works in other parts of their portfolio to manage those risks. Just over 50% of my portfolio is indexed. I happen to find it very effective. YMMV.

As part of supporting evidence for my claim, and to follow up on your challenge to me some time ago to go quantify my own personal rate of cost increases and ensure my portfolio is properly managed against that risk... (and now that 2009 completed and I had a chance to pull it together) I present both that result (my "real" return) and my investment results - at least 50% dependent upon indexing - by year since 1997 and by asset class - for your evaluation.

A few caveats and notes on the data:
1) Personal cost increase is measured from 2000-2009. Portfolio returns are shown for 1997-2009. The difference is primarily due to significant life changes that make it meaningless to bridge personal price increases before 2000. I do not have sufficient paperwork surviving to calculate portfolio returns prior to 1997, though I suspect it's immaterial to long-term IRR anyway.
2) Due to refinancing in the interim, I have "normalized" the home-cost data by assuming I purchased in 2000 and 2009 at then-current interest rates and then-current home assessment (conveniently assessed in 1999 and 2008) on the same duration mortgage. This raises 2009 costs (thus raising my personal price increase rate) over what I have actually experienced, but I believe is a more reasonable estimate of price increases. As an alternative, removing the mortgage altogether has a meaningless effect on the resulting personal cost increase from 2000-2009.
3) Due to the earlier period having a vehicle loan and my current vehicle fully owned in 2009, I have taken the 2000 model and charged 1/3 of the MSRP to 2000, and 1/3 of the same model at 2009 MSRP to 2009. My current model car didn't exist in 2000, so I used the 2000 vehicle to get an apples-apples comparison (This also raises my personal price increase rate, but is more accurate to the reality of having to replace vehicles)
4) For the financial returns I have removed my employee stock. Over the long term, it provides a minor (0.25% ish) positive delta to the portfolio IRR, but in individual years, especially years with a small portfolio, the purchase discount produces huge distortions (and I'm not willing to redo all the calculations by altering the data in Quicken or by doing it by hand in a spreadsheet). At this time, I hold no company stock and am not actively purchasing it.
5) For the portfolio returns, this does not include my pension. I simply do not have transaction-level detail in Quicken. It effectively is a medium duration bond that conveniently returns my personal price increase rate (zero real rate), and would represent approximately 9% of my portfolio at this time and shrinking on a relative basis because it is not receiving new money.
6) For asset class measurements, it does not include my 401(k) because that is in a fund that rebalances monthly (including worldwide stocks, bonds, real estate, commodities) and I don't receive IRR data on the underlying assets. I count Mutual funds under the appropriate class of the underlying contents of the fund, and I consider a CD a cash equivalent. I count all treasuries (bonds, bills, notes) as Bonds, which is probably inconsistent given that I call CD's cash equivalent, but I'm too lazy to fix it. If a stock was part of a covered call (or other similar transaction), both the stock and the call are counted as part of "Other Assets".

Over the time period from 1/1/2000 to 12/31/2009 my real rate of return (actual return over personal price increases) is a CAGR of 2.34%. This underruns my retirement model by 0.66% (Not bad for a "lost decade") and my early retirement model by 2.16% (so much for early retirement)

The returns stated below are IRR in nominal terms and do not count cash infusions into the accounts as "return". For reference of magnitude, based on Charlie's "Poverty-Risk Estimator" and using a real rate of salary increase of -1% (yes, that's a decrease in real terms) and the other data here, based on my current portfolio and rate of savings, I can live to 125.

Year    S&P 500   Me
1997 33.50% 22.80%
1998 28.66% 8.10%
1999 20.36% 33.10%
2000 -9.78% -10.40%
2001 -12.06% -9.70%
2002 -21.59% -7.20%
2003 28.17% 25.40%
2004 10.70% 15.60%
2005 4.83% 12.80%
2006 15.20% 14.00%
2007 5.15% 10.60%
2008 -36.79% -29.40%
2009 26.35% 31.80%

Currently ~13% of my portfolio is cash or short duration (<3 month) instruments, the vast majority of that raised since Thanksgiving 2009. About 33% is in bonds or bond funds. About 60% of my bond portfolio is effectively a "total bond market", 99% of the rest is investment grade, though I find Charlie's results playing with non-investment grade bonds to be interesting enough that getting an account at Zions or e-trade is in my plans for 2010, so I can at least evaluate what's available.

If I look at IRR of individual asset classes from 2000-2009 outside of my 401(k), I get this (nominal):
Cash/Cash Equivalent: 3.0%
Bonds: 14.2%
Large Cap Stocks: (2.3)%
Small Cap Stocks: 4.9%
International Stocks: 7.3%
Other Assets (options, etc.): 111.4%

As far as the recent crisis, while Charlie nibbled, my method and calculations showed the market was significantly distorted and so I wasn't as cautious. These are all more than single bonds. Then again, you will see I purchased mostly assets in the 10 year range, which is a significantly shorter duration than Charlie's nibbling, so I have much less capital gain.

GE Capital 36966RW28
International Paper 460146CA9
Prudential 74432AD98
MBIA 55252CAH3
Genworth 37247DAK2

These bonds were purchased in a ratio of roughly 5:5:5:1:1 (at purchase cost, not par). None represent what would be a concentrated position (>5% dependence on any one company) in my portfolio. 36966RW28 was sold in late 2009. I'm already 25% "out" (counting my dividends as return) of the MBIA, and I'm in well below par, so when it blows up (and I'm actually of the opinion that it's "when", not "if"), hopefully I'll stay above water.

Print the post  


When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.