Mad Money Machine

The Quest for the Perfect Portfolio

Yes, You Can Supercharge Your Portfolio! 2008 by Ben Stein and Phil DeMuth

I admit an initial bias against wanting to like this book. After all, these are the guys who wrote Yes, You Can Time the Market! And I’ve come to the view that you can time the market about as well as you can time the roulette wheel. You have to be right twice each time you try to time the market, both when to leave and when to return. I didn’t read that book, but I think it is wrong anyway whatever it says. Therefore I viewed this latest book by these guys skeptically. So to keep my hands clean and my bookshelf guilt-free, I borrowed it from the library instead of buying a copy.

My skepticism turned quickly to excitement once I got inside. These guys are talking my current passion, Lazy Portfolios and buying the best basket of index funds you can assemble. They are talking return-to-risk ratios. They are talking standard deviations. They are talking uncorrelated asset classes. These guys are going to give me the answer to my question, "What is the Perfect Portfolio?" I flipped pages quickly, impatient to get the answer. Hey, I thought, I may want to actually buy a copy of this book. I want to keep a permanent record of the answer.

I got an answer alright. But it wasn’t what you might expect. But let’s start at the beginning and set the stage before I give you the answer. The authors lay this portfolio supercharging process out in what they subtitle as Six Steps for Investing Success in the 21st Century. The first few steps start typically. Know what you are buying. Diversify. Know your risk. Diversify. And so forth. Some nuggets embedded within are particularly appropriate to lazy portfolio investing. Such as the notion of dividing the annualized historic return by the annualized standard deviation resulting in the return-to-risk ratio. On my Lazy Portfolio graphs, those portfolios with a high return-to-risk ratio would be plotted at the top left of the chart just where you want to be.

What the authors never told me though is that there is another well-known return-to-risk ratio measurement called the Sharpe Ratio. And they never told me that the Sharpe Ratio can be easily found at finance.yahoo.com for most any mutual fund ticker you can type, such as for VGPMX for example. Instead, they told me to go and buy a program called Quantext Portfolio Planner (QPP) and calculate some funny things myself. I’ll get to this in a second. Let’s return to my excitement.

You’ve seen that I’ve been tracking a number of Lazy Portfolios by professional investing gurus. So naturally when I saw Table 4.3 listing the Couch Potato, Margarita, Six Ways from Sunday, Diehard, Coward, and Coffeehouse portfolios, I knew it would be a late night staying up with this book to get to the answer. And this table confirmed what I have also witnessed, that the Six Ways from Sunday portfolio developed by Scott Burns has outperformed all others and by a nice margin. How exciting it was to see that in the very next chapter the authors take this Six Ways portfolio and Supercharge it using a Monte Carlo simulator. Wow, do even better than the best portfolio? Let me call my broker just as soon as I get the answer.

How can we do better? They assert that we can find higher return-to-risk portfolios by adding in investments that are non-correlated and that themselves have high return-to-risk ratios. So when you’ve run out of non-correlating, high return, low risk index funds, where do you turn? Shockingly, the authors suggest that we turn to individual stocks. So help me Cramer that’s what they said. But first we need tools. Tools will help us get the answer.

Right. So after detailing the benefits of diversification and non-correlated asset classes, the authors take us into the land of the Monte Carlo simulation. (How appropriate that in the real Monte Carlo they have roulette tables.) This is the point where the authors introduce us to the QPP Monte Carlo simulator which is basically an Excel spreadsheet for $85 per year that pulls historical fund quotes from Yahoo! Finance and runs some stats and simulations against them. Hey, haven’t I heard of something like this before? Oh yeah, the Stock Market Functions Add-In and the RCHGetYahooHistory function that I use to calculate annualized standard deviations on portfolios in Excel. But of course their tool does much more than standard deviations — they do Monte Carlo simulations! Which are what, exactly? The authors say a Monte Carlo simulator:

uses a computer’s random number generator to construct sample sequences of future returns. After spawning thousands of possible scenarios for a given portfolio, the shape of its financial future begins to take form.

Most important, Monte Carlo simulation can put different portfolios through the same set of paces, letting us make head-to-head comparisons… It’s an effective tool to sharpen investment decisions.

Basically, the simulator looks at the range of past results for each fund and then rolls a matching set of dice thousands of times to see the range of possibilities that could happen to the whole portfolio in the future. It’s magical! I guess, since the authors don’t show any supporting facts that having run Monte Carlo simulations in the past have actually resulted in good things happening in the future. Let’s just believe that we can run this QPP and get a better idea of success than by just simply looking at past Sharpe Ratios, OK?

But what if your range of historical past results is too short? For example, let’s say you are trying to figure out the range of possible values of five dice rolled together by rolling them many times. These guys effectively rolled the five dice just 36 times. Dudes, you’re going to need a lot more than that to predict the future possibilities. That is one flaw in what the authors are showing here. They used only three years of historic data to show returns and standard deviations for assets. And even worse, the three years they chose were from 2003 to 2006, a major bull market in stocks. Any monkey would make money by buying stocks in 2003, so come on.

A further flaw: they don’t elaborate on exactly what percentage of risk, what standard deviation of a portfolio, that one should buy into. They do say that "risk is not psychological" and that we should not be talking about risk tolerance, we should be talking about "running out of money tolerance." But that is basically where they leave it. For listeners to my show and readers of this blog, you’ll know that the first step is to take the Risk Capacity Survey at IFA.com. It will give you a better idea of what risk level you should seek. Then Index Funds Advisors can help you build a portfolio to match that risk level.

But the biggest flaw of all is in back-testing individual stocks to try to find ones that will do better than the market. They think that by adding 10% individual stock picks you can make your portfolio have a higher return with a lower risk. Keep dreaming guys. I’d guess that of the folks that try this, that 45% will beat the market and that would be just by sheer chance alone.  Kind of the same success rate of picking Red to win on the roulette wheel, wouldn’t you say? And that margin of "beating the market" is so slim that it would have been a much wiser use of one’s time to sell one’s old junk on Ebay or something.

But oh what a price to pay to try to achieve that slim margin. First, you have to buy their book. Second, you have to buy this spreadsheet tool. Then get it working. (I obtained a trial copy. It did not work immediately. It did not work after I tried a couple of things. I did not pursue it further. It must work for someone because the authors report its results extensively.) Third, you have to spend time guessing which stocks to pick. Then you have to buy the stocks. Then you have to monitor the stocks. Then you panic and wish you had bought different stocks.

Finally you come to the realization that just sticking purely with index funds is the right way to go. Which brings us to the answer: No, you cannot supercharge your portfolio! 

Here are some other relevant quotes to consider when deciding whether to use 3 years of data to create a Monte Carlo simulation:

"Statisticians will tell you that you need 20 years worth of data — that’s right, two full decades — to draw statistically meaningful conclusions [about mutual funds]. Anything less, they say, and you have little to hang your hat on. But here’s the problem for fund investors: After 20 successful years of managing a mutual fund, most managers are ready to retire. In fact, only 22 U.S. stock funds have had the same manager on board for at least two decades–and I wouldn’t call all the managers in that bunch skilled. "
by Susan Dziubinski, University editor with Morningstar.com
Note: Index Funds are the only source of reliable 20 year risk and return data.

And since picking individual stocks essentially turns you into a portfolio manager:

"Studies show either that most managers cannot outperform passive strategies, or that if there is a margin of superiority, it is small." p. 372-b.

"It will take Joe Dart’s entire working career [calculated to be 32 years] to get to the point where statistics will confirm his true ability." p. 821 -  c.

"In the end, it is likely that the margin of superiority that any professional manager can add is so slight that the statistician will not easily be able to detect it. " p. 374

Zvi Bodie, Alex Kane, Alan J. Marcus, Investments, Fifth Edition, McGraw-Hill (Thanks Mark Hebner for the quotes.)

I feel the sudden urge to write a book entitled, The Quest for the Perfect Portfolio in which I will compare all of the Lazy Portfolios and try to construct THE ONE portfolio appropriate for everyone by simply adjusting the percentage of bonds vs. stocks. Oh, wait a minute, Mark Hebner has already done that: Index Funds: The 12-Step Program for Active Investors.

Feel free to read comments at the blog.