The Canadian Source Of Employee Pension Fund Investment And Benefits Plan Management

Back Issues

Risk: The Good, The Bad, And The Ugly

By: Paul Dontigny Jr. & Eric Fontaine

The quest to reduce risk, driven in part by the legal system, has resulted in investors forgetting that the primary purpose of financial markets is to allocate capital to maximize wealth. Paul Dontigny Jr., of Investissements PDJ, and Eric Fontaine, of Brockhouse & Cooper Inc., examine how this occurred and its impact today.

When Eric Fontaine and I initially discussed writing an article on portfolio management, it was because we agreed on a few major issues concerning institutional investing and their combined potential impact on financial markets. The issues are mainly related to the practices of relative investing, indexing, and benchmarking. But an interesting thing happened as we tried to focus on the subject and text … something that illustrates the practical complexity of the problems facing pension funds. We did agree on identifying the major problems and challenges, but even then, we realized we did not have the same perspective.

Eric’s view is influenced by his actuarial and investment background and based on his experience as a pension/asset consultant with pension funds. His perspective is a pragmatic one, trying to help his clients improve the risk-reward characteristics of their pension plans, taking into account the assets and liabilities as well as the company’s risk profile.

My view is that of a portfolio manager free to invest long or short in any asset class, in any currency, with the objective of maximizing positive absolute returns over five year periods. My perspective for this article, because pension assets represent a substantial portion of total stock market capitalization and a substantial expense to corporations, is to try and figure out the impact of major shifts in pension investment style on stock and bond market valuations.

A Legal And Political History Of Risk

By: Paul Dontigny Jr

Peter Bernstein has recently shaken common institutional wisdom on the subjects of indexing, benchmarking, and long-only strategies. He also cast a doubt on the validity of investment policies that basically fix asset allocation in tight ranges (about 65 to 70 per cent stocks in the U.S.) for long-term investment.

In that context, I will provide some historical perspective on what I believe has led us to where we are now. I will also question the sustainability of current pension investing methods, taking into account the purpose of markets and investing.

So let me start by reiterating the position I presented in the AIMR Exchange (September/ October 1999): Modern portfolio theory (MPT) has created a bubble because of misunderstanding and misuse of otherwise fundamentally sound investment principles. It is ultimately from this theory that has evolved, over 50 years, an investment business focused on short-term quarterly comparative results with peers and popular indices.

The U.S. legal system has had as much to do with major investment industry developments as investment principles and theories. And, in the quest for reduced risk and legal protection of investors based on quantifiable investment principles, we seem to have forgotten the purpose of financial markets. The central purpose of the stock market is to allocate capital in a way that will maximize wealth in the economy over the long run. It’s not to allocate risk in order to minimize quarterly volatility of a portfolio as many seem to have concluded from Markowitz’s and Sharpe’s theories.

The Evolution Of Prudence

On the legal side, it all started with the ‘Prudent Man Rule,’ enunciated by a Boston judge in 1829. The interpretation of this ruling has evolved, but for many years, starting in the early 1900s, the courts applied this rule in a way that forced any investor in a fiduciary position to avoid risk of any capital loss. So in the U.S., most insurance companies and pension funds up, until the mid-1940s, invested almost exclusively in fixed income and in assets enumerated in the ‘Legal list statutes.’ This list, for years, permitted only investment in major financial institutions and very few major corporations.

With the still fresh memories of 1929, 1932, and 1937, when the stock market plunged about 50 per cent within one year, that type of prudence was understandable. It is only in the early 1950s that stocks started to be a larger part of U.S. pension assets. It has been suggested1 that the main reason may have been that major corporations like GM and US Steel wanted to have a say in pension asset management as pensions were becoming a significant aspect of collective bargaining.

As of then, with the strong bull market and inflationary pressures from the Korean War, a ‘Man’ (or a pension fund) could be deemed ‘Prudent’ if he purchased stocks of good companies. New York laws were also changed to allow Legal Trusts to buy stocks in limited amounts.

And as human nature expressed itself for a while taking the market ever higher into the ’60s, I take you to the fourth annual Institutional Investor’s Conference in March of 1971, where David Babson answered the question ‘what went wrong in the late 1960s?’ He compared that situation to the Mississippi Bubble and the Dutch Tulip Craze. Babson listed 12 things that went wrong including, to my surprise, “the modern corporate treasurers who looked upon their company pension funds as new-found ‘profit centres’ and pressured their investment advisers into speculating with them.”

And from these excesses was born the ERISAAct of 1974.

ERISA was built in a time of stock market and economic crisis, but also in a revolutionary time of financial academic research.

It is in that context that risk was redefined as being not the characteristic of a security, but that of a portfolio. The important point here is that this new definition was simultaneously put into effect by law. It is automatic from this change that trustees and portfolio managers would then, by law, be subject to short-term monitoring of performance relative to benchmarks.

At the same time, numerous studies seemed to confirm that indexing is superior to stock picking and that a fixed, constant asset allocation is superior, less risky, than the so-called market timing of different asset classes. And now we even hear that it is more risky NOT to own stocks than to own 55 per cent of them in a portfolio. The Modern Prudent Investor has to own stocks at all times, period.

RISK: Absolute Value Of The Benchmark Itself

Being legally responsible to these principles has, of course, proven more powerful than being ideologically convinced that buying good assets at good prices based on subjective valuation is a good approach. And that is how we ended up today with more than a third of the market’s capitalization being invested on a relative basis. The result is that no one in the pension investment process currently has the responsibility, the desire, or the power to question the absolute value of the benchmark itself.

Nowadays, risk is commonly defined as tracking error to an index or short-term portfolio volatility. In this environment, it is an implicit assumption that for long-term investments, stocks are superior to bonds for purchase at any point in time, independently of price. Under this false assumption, the stock market is clearly not fulfilling its purpose of allocating capital based on economic value and maximizing wealth. And risk seems to be defined more in terms of shortterm risk and job risk, than in terms of the probability of fulfilling the long-term obligations of the pension fund/pension plan.

Because of the extent of these common wisdom beliefs about risk, it is important to remember that Markowitz’s quantification of risk was based on the standard deviation of expected dividends (fundamental analysis and absolute valuation). Stock price volatility (technical analysis) was initially used only as a proxy in the required empirical study to complete his article.

Sharpe later used stock volatility in his equilibrium model because of high computer costs at the time, and unavailability of data for fundamental analysis and forecasts. But with current data and technology available, any true measure of risk must relate to the capacity of a company to earn future profits (or a bonds’ capacity to earn interest and capital) and the investment value must so be determined to calculate the expected return and risk. Taking these into account, pension fund investment policies need to revisit the definition of risk, which some committees have already started.

Risk is not only volatility and volatility is not necessarily bad.

Two Types Of Volatility

By: Eric Fontaine

I will attack the current perception of risk by presenting two types of volatility, the good and the bad, and use two examples to show the benefits of high volatility for a pension plan.

It is a well-known fact that pension plans are long-term liabilities and, as a result, the assets supporting them should be invested with a long time horizon.

What is less well known (or less well remembered) is that from the late 1950s to early 1970s, most plans were invested in insured contracts, managed by insurance companies, and that the pension liabilities were almost entirely invested in Fixed Income instruments under immunized strategies (with limited investment risk).

Then, in the late 1970s and early 1980s, independent money managers emerged and started offering plan sponsors active equity management that led to the creation of Balanced funds or mandates. This was a major turning point which saw the Management community (the Optimists) take over from the Actuarial community (the Conservatives).

It was the moment when the industry stopped thinking about liabilities and started focusing on the asset side of the equation and the concept of mismatch-risk was introduced.

Bearing in mind the long term nature of pension plans, the general perception was that the mismatch-risk would be more than compensated for by additional portfolio returns generated by the introduction of an equity component. But something else changed.

Benchmark Risk For Bonds

Prior to the 1980s, the most common Fixed Income benchmark used by the broad pension community was the Long Term Canadas. The Scotia Capital Market Universe (SCMU) – today’s most common Canadian bond index – was only introduced by Scotia McLeod in ....1980! Was this pure coincidence or were the ‘Optimists’ taking over?

Since the Management community likes to talk in terms of broad Universes (TSE 300, S&P 500), the SCMU, which displays characteristics similar to mid-term bonds,2 became the obvious choice.

However, in doing so, another mismatch- risk was created, the duration-mismatch. Again, because liabilities were excluded from the risk analysis, longer bonds, because they exhibit, by nature, higher volatility, were perceived as riskier than shorter bonds.

However, when the liabilities are taken into account, it tends to have the reverse effect. The higher bond volatility is neutralized by the long-term nature of the liabilities (similar duration characteristic). In a nutshell, since both the assets and the liabilities move in tandem, it becomes less risky to invest in longer term bonds. This is what I call positive or ‘Good Volatility.’

Another important turning point also occurred in the late 1980s when major changes were made to the federal and provincial legislation affecting pension plans. Following these changes, the dynamic of pension plans became much more short-term in focus as a response to solvency/wind-up requirements.3 The move to a shorter term focus became even stronger when pension accounting standards4 were introduced in the early 1990s (late 1980s in the U.S.). In recent years, many articles have been written to educate investors on the importance of the pension accounting figures and the significant effect they can have on the Balance Sheets (shareholders’ equity) and the Income Statements (Earnings and EPS) of public corporations. It goes without saying that the latter are exceptionally short-term focused.

Downside Risk For Equities

In light of these changes, you will appreciate that, although pension plans are long-term in nature, many short-term issues come into play. For this reason, pension plan fiduciaries should also pay more attention to ‘downside risk’ (referred to as Bad Volatility) and not just the overall equity risk which is, typically, measured by Standard Deviation.

Earlier in the article, I presented the case where a longer term bond portfolio provides better downside protection than a broad benchmark that bears little relation to the liabilities.5

To show the pitfalls of Standard Deviation as a measure of risk, let us analyse two equity managers showing the following historical return pattern and characteristics – CR Compounded Returns; SD Standard Deviation; SR Sharpe Ratio).

Looking at Chart 1, Manager A, using the Sharpe Ratio, seems to have outperformed Manager B on a risk-adjusted basis. However, since the Standard Deviation does not differentiate between upside (good) and downside (bad) volatility, both the standard deviation and the Sharpe ratio favour Manager Adespite the fact the Manager B has outperformed the other by two per cent per annum. In fact, we note that over the period under study, Manager B outperformed both the benchmark and Manager Ain each and every year, and consequently did not display any relative downside risk. This signifies that the historical volatility has always been positive (again here I refer to the Good Volatility). Depending on the objectives (focus), downside analysis can be performed at either the absolute or relative level.

What I have tried to highlight with these two examples, is the difference between good and bad volatility as it applies to bonds and stocks. I have also noted the importance of having a good understanding of the plan sponsor’s liabilities. Even at today’s level, longer term bond benchmarks (in real or nominal terms) may still represent the best option for pension plans.

Like many over the past few years, I hear you say, “But what if yields go up”? Truly, I don’t know where rates are going (plan sponsors hire and pay good money to managers to deal with this issue).

What I do know is that if rates are going up, it will mean that your bond portfolio is going down, and so will your (marked-tomarket) liabilities.

The real question you should be asking is “But if the yields continue to go down, what would the downside risk be and what impact would it have on my plan’s financial status?

The Final Word

The common conclusions to our perspectives on the challenges facing pension funds are quite simple and, in our view, reflect economic common sense. But the challenges are politically and legally very difficult to solve. Solving them might also create instability in financial markets if, for example, the result of policy revision was to decrease substantially the asset allocation into stocks.

We are conscious that we write these words in a period of fragile world economies, volatile financial markets, and substantially underfunded pension plans, some of which are larger than the companies that fund them.

That is the ugly … and there are clearly no easy quick fixes.

The importance of the pension plan/fund issues cannot be underestimated. Because of their size, pension funds directly and materially affect corporate profits and employee’s life savings and retirement lifestyles, as well as the stability and valuation of financial markets. We believe that the investment and legal notion of risk should, and will, be redefined, as it often was in the past. This time, changes will be made to the current notion of portfolio benchmarking so that pension investment in assets will take into account the dimensions of long-term absolute returns and pension liabilities.

Investment policy should not restrict asset allocation solely based on long-term average historical returns. Rather it should promote the use of fundamental economic valuation of each individual asset or security. Valuation of benchmarks might become another trend that would lead to more dynamic asset allocation changes based on economic value, not on price trends.

When ERISA was enacted, pension funds were not as big a proportion of total market capitalization as they are today. Because they are such significant players in the market, the governing laws and regulation of pension fund must be built so that they allow and support the markets to serve their purpose of allocating resources in the economy to more productive and profitable uses, and not to overvalued stocks or bonds.

With the Bernsteins of this world now being vocal about these principles, and with the ongoing review of regulation in the industry, it seems the trend will be towards a return to individual company and bond fundamental analysis and valuation. We could call it a bottom up asset allocation based on economic value estimates.

And diversification of risk will relate, as it should, to economic factors instead of relating to historical price charts. These are very positive long-term developments for the markets and the economy, even though it might hurt substantially in the short term.

Paul Dontigny Jr. is president of Investissements PDJ (pdj@videotron.ca). Eric Fontaine is a senior consultant at Brockhouse & Cooper Inc.

1. The birth of a trust department, reprinted in Classics II (Charles Ellis) from “some comments about the Morgan bank” (1979)

2. These days, the duration of the SCMU is around six years compared to 11 years for the long Universe. But the duration of a typical pension plan might range from 10 to 15 years.

3. In Canada, most federal and provincial legislation require that any solvency deficiency be funded over a period not exceeding five years. The solvency assets and liabilities are marked-to-market at the date of determination on the assumption the plan will be totally terminated.

4. The accounting actuarial valuation represents a mix of solvency and funding characteristics as the liabilities are projected (on-going like funding) and both the assets and liabilities are marked-to-market (like solvency).

5. Note that in the U.S., the duration mismatch has been even larger than in Canada, taking into account that the most common broad universe indices have displayed much shorter duration than the SCMU.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Subscribe to Daily News Alerts

Subscribe now to receive industry news delivered to your inbox every business day.

Interactive issue now onlineSubscribe to our magazinePrivate Wealth Online