Do you enjoy a wild ride? (photo credit)

Do you enjoy a wild ride? (photo credit)

Asset Allocation is a fundamental part of investing.  Which investments should we own?  What percentage of total investments should we allocate to equities, bonds, Real Estate, or alternatives?  As an American, should I purchase International assets, or focus solely on US assets?

There is an incredible amount of research on this topic, considering factors from expected duration of retirement to personal temperament.  Perhaps you’ve heard some of the sound bite versions of this research, such as “Hold a percentage of stocks equal to 100 minus your Age.”  So if, for example, you were 40 years old, you would hold 60% stock and 40% bonds

Like most mainstream investment advice, it is targeted at people that plan to retire at 65 and live until they are 80.  Which is why as an early retiree, this advice (and most mainstream advice) is harmful at best

Tossing conventional wisdom to the side, over the past several years I’ve been moving our asset allocation towards 100% equities.  Let’s explore why

Risk

Wait a second?!  Isn’t 100% stock super risky?  What if the stock market suffers a major drop?  What if the Great Recession happens all over again?

I think the phrasing of these questions holds an underlying assumption, that the stock price on any given day is important.  Unless you plan to buy or sell, the price is largely irrelevant.  I seldom look at the stock market.  The last time I did so was to participate in Jim Collins’ annual Market Prediction Contest, a competition I predict I will lose 3 years running

100% guaranteed, over our planned 60+ year retirement the stock market will have many major declines, several of them will be sustained over long periods.

But this does not concern me.  Our real risk comes not from changes in share price, but from outliving our portfolio

The Case for 100% Stocks

Asset Allocation and Success Rates

There are several different methods to predict portfolio longevity, or retirement success rates.  The two most common methods are probably Monte Carlo analysis and Sequence of Return simulation, both of which can be done with cFIREsim.

For a typical 30 year retirement, we can use cFIREsim to predict portfolio success rate for varying stock/bond allocation

I use all default values (4% withdrawal rate) for “Success Rates with Various Allocations”, with the exception of Fees.  For this, I use the Personal Capital Retirement Fee Analyzer which tells me I’m paying 0.08% with current asset allocation.  If you don’t know your current investment fees, this tool will figure it out in a few seconds

Success Rate for 30 year Retirement. Peak success at 90% equities

30 year Retirement Success Rates. Peak Success at 90% equities

Generally speaking, this data suggests any asset allocation from 60-100% equities has about the same chance of success (90%+.)  I more or less assume anything above 80% success rate is false confidence.  The future will likely have many Black Swans. You never know if Tyler Durden is going to erase the debt record, Simian flu will wipe out 90% of the human race, or a terrorist will detonate a dirty bomb in downtown Manhattan.

There are fewer 60 year investment returns to explore, but doing so gives a slightly different picture

60 years

60 Year Retirement Success Rates.  Peak Success at 90% Equities

Again the peak success rate is in the 90% range.  The success rate with a high percentage of bonds looks quite dismal.

For completeness, let’s also look at a Monte Carlo projection.  This uses statistical models for stock and bond performance, and simulates thousands of retirement periods to come up with a confidence factor.  I don’t favor this method, as data for each year is random which ignores correlation between years as well as the fact that mean and standard deviation should change based on economic environment.  Also, in the cFIREsim model annual inflation is fixed at 3%, which ignores periods of deflation or high inflation

Because of the random nature of the analysis, the results will never be the same twice.  But here are 2 outputs, one for 30 years and another for 60 years

30 yeras

Monte Carlo Analysis Success Rates for 30 Year Retirement

60 years

Monte Carlo Analysis Success Rate for 60 Year Retirement

In both cases, holding 70%+ seems to produce success rates in the 90% range

Return vs Volatility

The Efficient Frontier is often used to graphically represent return vs volatility for different asset allocations.

I pulled this analysis from the Personal Capital Investment Checkup tool.  Based on our current (not 100% stock) asset allocation, we are looking at a potential average return of 9% with a standard deviation of 14.3%.  By moving to 100% stock, we have a potential annual return of 10.1%, but with a substantially greater volatility of 20.2%.

Most people will argue that a small increase in annual return is not worth the vastly increased volatility.  This is especially true when the success rate of our portfolio over a 60 year period is roughly the same whether we have 70% equities or 100% equities

So why accept the greater volatility?  Because over a 60 year period, even a small increase in return results in massive differences in total assets

Efficient Frontier

Efficient Frontier

Terminal Value

In addition to evaluating success rates, cFIREsim’s Standard Simulation can help us estimate portfolio value statistics

Using default values except for our own fee structure, we get the following results for equity percentages of 70% to 100%.  All data is for a 30 year retirement, with two results for a 60 year retirement shown in parenthesis

% Equities70%80%90%100%
Success Rate94%94%94%94%
Ending Value
Average
$1.8$2.1$2.4$2.8
Ending Value
Median
$1.2
($3.0)
$1.6$1.9$2.3
($12.3)
Ending Value
Std Deviation
$1.5$1.7$2.0$2.3
Ending Value
Highest
$5.6$6.4$7.2$8.8
Ending Value
Lowest
-$0.3-$0.3-$0.3-$0.3
Cycles dipped >60% below initial value28 of 115 (24%)27 of 115 (23%)27 of 115 (23%)24 of 115 (21%)

The difference in median value for a 30 year period over the studied range is 2x.  Over a 60 year period, it is 4x.  And this is without a major difference in success rate, cases with dips of greater than 60% in value, or minimum terminal value

Posterity will benefit tremendously from our tolerance for risk

Arguments Against 100% Stocks

The numbers say 100% stocks is the right way to go.  I agree

But what if future market returns are not as strong as the past?  And what about emotions?  Humans are not logical machines.  How many people panicked and sold at the bottom in 2008, losing everything?

There is no one right answer.  I am optimistic about future economic performance, and am comfortable with volatility, which is why I can reminisce about the glory days of 2008.

Tolerance to Risk

If somebody is not comfortable seeing the value of their portfolio collapse, having 100% of their investments in stock is not a great idea

We lost 10 years worth of spending in 2008.  Some people freaked out.  I did too, but for a different reason.  I was upset that I didn’t have extra cash to buy more stock.  During the downturn, I realized the value of the portfolio didn’t impact our life one little bit.  Dividends were still being paid, and our daily lives were exactly the same

In recent times, I’ve watched the value of our portfolio swing by several years worth of spending.  Dividends were still being paid, and our lives were exactly the same

If the stock market were to collapse tomorrow, dividends would still continue to be paid and our lives would be exactly the same.  And I probably wouldn’t know about it for a few months anyway, since I don’t often look at the market and don’t watch television

Margin of Safety

We also have incredible margins of safety that would prevent us from selling stock in a downturn.  In the big picture, most of these would only need to be used in the next 10 years or so.  At some point, our portfolio will become too big to fail.  cFIREsim already predicts a 100% success rate without future income or Social Security

This is largely because nearly 100% of our over the top luxurious lifestyle is funded by dividends and interest.  Dividends can be reduced or eliminated, but we can also reduce our spending (In 2008, VTI/VTSAX dropped the dividend 3%.  In 2009, a further 12%.  It has more than fully recovered)

In a fixed location, we can reduce spending by eating more tofu and less steak and lobster.  Or we can practice geographic arbitrage, eating steak and lobster in Belize or Argentina instead of Paris or Tokyo or Sydney.  Being 100% location independent has its privileges

Some would argue that there is no need to continue to seek growth.  Perhaps I agree, despite the clear mathematical advantage of 100% Equities.  Which is probably why our current asset allocation is only on the path to 100% equities rather than having arrived

The GCC Asset Allocation

Using the Personal Capital Portfolio View, I get the following snapshot of our portfolio

Screen Shot 2015-03-04 at 9.15.04 PMUS Stocks is 95% VTI.  International Stocks is 100% VEU.

US Bonds are a roughly equal mix of BND / IEI / MUB / TIP

Alternatives is 100% VNQ

Cash is our Capital One 360 Account

Update: our portfolio moved closer to 100% equities in 2016.

What is your Asset Allocation plan?  How do you feel about 100% Equities?