No. of Recommendations: 4
I have been struggling for some time to come up with a better terminology for the two types of factors that I see as affecting long-term stock returns. My usual practice has been to refer to the factors focused on in the conventional methodology SWR studies as "volatility" and to refer to the factor not given adequate consideration as "valuation." I have never been happy with these terms. I have been tempted at times to use the terms used by William Bernstein in Chapter Two of “The Four Pillars of Investing”; he uses "risk" for the factors focused on in the conventional methodology studies, and "return" for the factor that I have been stressing. I was recently reading through some old posts in an effort to gain a better understanding of the definitional issues, and I have come up with a new terminology which I think may clarify the distinction that I have been trying to put forward.

My new terminology is to refer to the factors focused on in the conventional methodology studies as "Imposed Risk" or "Unavoidable Risk" and the factor not given proper consideration in the conventional methodology studies as "Voluntary Risk" or "Elective Risk." There is a sense in which the latter types of risks do not constitute true risks at all. Bernstein says on Page 12 of “Four Pillars” that “where there is return, there also lurks risk.” I believe that that is almost correct, but not quite. There is an important sense in which stock investors can tap into higher long-term returns by incurring Voluntary Risk or Elective Risk without incurring any additional real risk. Much of the risk you take on when you purchase stocks at times of low valuation levels can be said to be Risk without Risk.

SWR analysis is a risk-assessment tool. The purpose is to use historical data to assess what sort of risk you are taking on in investing in stocks to finance your retirement. The conventional methodology aims to do this by determining the worst return scenario that has turned up in the historical record and declaring that a withdrawal strategy that survived that scenario is safe presuming that the future is no worse than the past.

The core flaw is that the conventional methodology fails to make a distinction between two types of risk with very different characteristics. There are many factors that cause stock investing to be risky. Demographics changes cause risk. The rise of the country you are investing in as a world power, or its fall from that position, causes risk. Productivity changes cause risk. Risk results from the fact that the future is unknown, and there are many aspects of the future that are unknown. So there are many things causing stock investing to be risky.

The problem with assessing the risks you face as a stock investor is that there are so many different aspects of the risk question that it is impossible to sort through them all. You might feel confident that demographics issues are going to be worse in the next 30 years than they have generally been in the past. But you might feel that productivity issues are going to be better, and that the latter effect will more than cancel out the former. So many factors come into play that it is impossible to analyze them all separately in a satisfactory way.

SWR analysis comes to the rescue. SWR analysis combines all of the various risk factors into a single number. It doesn't make a claim that demographics will be worse or that productivity will be better or anything else. It identifies what is the worst COMBINED effect that we have ever seen in this country, and declares a withdrawal strategy that works under those conditions "safe."

I view this as generally being a good way to proceed. SWR analysis reduces all of the many risk factors down to a single number and thereby helps the investor develop his own personal withdrawal rate (PWR) in an informed manner. Each investor makes adjustments for his own optimism or pessimism by making adjustments to a SWR number that reflects one optimistic assumption ( a future not worse than the past) and one pessimistic assumption (a worst-case scenario popping up in one's retirement).

My objection to the conventional methodology is that it treats the valuation factor in the same manner as all the other factors. It is one more risk factor thrown into the mix; no separate adjustment is made for changes in valuation. Why is it that I believe that doing that is an analytically invalid procedure for a tool aiming to inform investors as to the level of risk they are taking on by using stocks to finance their retirements?

It is because the nature of the risk taken on by investing at high valuation levels is of a fundamentally different nature than the risk taken on by leaving your retirement subject to ever-changing productivity and demographic rates. The core difference is that the risk taken on by investing at a specified valuation level is known at the time the stocks are purchased. All of the other risk factors are Imposed Risks. The valuation factor is Elective Risk. Each investor has a choice as to whether to take on the risks involved in investing at high valuation levels or to not do so.

This is the thing that most bugs me about the conventional methodology. I love the idea of packaging all the various risk factors in a single number so that I can make informed investment decisions. But it does not seem right to me to treat the elective risk component in the same manner as all of the unavoidable risk components. I can take an educated guess as to what is going to happen with demographics or productivity and so on, but that's all that I really am doing when I state an opinion as to the effect that these factors will have on my retirement plan. This is not so of the valuation factor. I can state with mathematical certainty that investing at times of lower valuation will bring me better investing results and know that there has never been a time in history when this was not so (all other factors being held equal).

I do not want my SWR tool treating valuation in the same way that it treats all the other risk factors. I want to use the tool to make decisions as to how much of my portfolio to allocate to stocks, and in order to use it for that purpose, I need it to differentiate between the Imposed Risk factors and the Elective Risk factor. There is a sense in which Elective Risk is not risk at all. At a minimum, it is a different sort of risk than Imposed Risk. To do the job we are asking it to do, the SWR tool needs to separate out the Elective Risk factor and show clearly the effects of the Imposed Risk factors. Since the effects of the two types of risks are mixed together in the historical data, the Elective Risk component must be treated separately for the SWR analysis to provide an informed answer to the most important question being explored--What amount of Imposed Risk am I taking on with this investment choice?

Say that you are buying a $1 lottery ticket, and the odds of winning a $50 prize are 1 in 100. You are taking on a risk, and the statement of odds tells you the extent of the risk you are taking. Knowing the odds tells you what you are getting into, somewhat akin to the way that SWR analysis aims to tell you the risk you take on with stock investing. The worst-case scenario with the lottery ticket purchase is a loss of $1; the best case is a profit of $49; and there is a 99 percent chance of the worst case coming up and a 1 percent chance of the best-case scenario coming up.

Now say that it is not always easy for you to get to the store and purchase a ticket. There is a guy at work who is willing to buy a ticket for you, but he charges you $1.20 for each ticket you purchase to compensate himself for the trouble he incurs doing so. If you buy the ticket from him, should you be recalculating the odds of a successful outcome? There's a sense in which you should and there is a sense in which you shouldn't. The extra 20 cents you pay definitely affects your return, so you definitely need to take it into account in making decisions about whether to purchase lottery tickets and how many to purchase. But it is not the same sort of expense as the one you incur when you pay the $1.00 purchase price for a ticket.

The extra 20 cents is properly classified as a convenience expense; you are incurring it to avoid having to take a drive to the store that sells the ticket. That expense is not a necessary component of a purchase of a lottery ticket, and it confuses your analysis of the risk involved in purchasing lottery tickets to treat it in the same manner as the purchase price amount.

I think that something similar is true in regard to stock purchases. There are some risks that are unavoidable, that simply must be incurred in all stock purchases. These are Imposed Risks. The level of risk incurred through these items is reasonably well conveyed by the conventional SWR analysis, in my view. But the extra expense that is incurred for purchasing shares at times of high valuation is not well conveyed. When you mix that factor in with all the others, you convey the impression that valuation is just one of many risk factors when it is really a special one, one that the investor can eliminate by electing to purchase shares at strategically advantageous times.

The other side of the story is that purchasing at low valuation levels allows the investor to minimize the price he pays for obtaining access to a given long-term return expectation. It's like having a friend offer to sell you his lottery ticket at a reduced price because he changed his mind about the wisdom of his purchase. In those circumstances, you are getting a discount on the normal purchase price. The discount aspect of the question needs to be taken into account in your decision as to whether to make the particular purchase. But it is not quite right to say that the discount you are offered changes the risk involved in the purchase of a lottery ticket. It changes the risk for that particular purchase. But if you start thinking that all lottery tickets can be purchased for 80 cents, you are going to develop an ill-informed understanding of the true risks involved in purchasing lottery tickets.

What I don't like about the conventional methodology is that it misleads people about the risks they are taking on when using stocks to finance their retirements. By putting forward the same number at all valuation levels, it suggests that valuation doesn't affect the result any more than any other factor. Many investors have come to believe that this is more or less true, that a purchase made at a high valuation level works out over time because stocks are always the best investment class in the long term. The reality is that it never works out to pay more for a specified level of long-term expected return than you can afford to pay. When you pay a higher price for a specified amount of return, you incur a permanent cost.

I prefer an SWR tool that makes that clear, that separates out the factor that is under the control of the investor from the factors that are not under the control of the investor. I think that it is misleading to treat the starting-point valuation as just one of many risk factors when in reality this is a very special risk factor, a risk factor under the control of the person making the investment. Conventional methodology studies provide valuable insights. But such studies always give the wrong answer to the ultimate question being examined--What is safe?

So I view the conventional methodology as an analytically invalid means of exploring the question of what is safe. I believe that there should be an adjustment for the risk factor that is incurred at the election of the investor. I believe that the question that investors really want an answer to is, What risks am I taking on that are outside of my control? It is not possible to provide a good answer to that question without treating the Elective Risk factor differently than the Imposed Risk factors.

Risk without Risk is special. It is the proverbial free lunch, money for nothing (except for the emotional strain associated with buying stocks when all of your friends are selling them instead). I hope that we will be exploring the implications of the Risk without Risk concept in much more depth in the months and years to come.
Print the post  


What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
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.