2020 Vision

Investment wisdom for tomorrow

Harry Liem

• Publisher: Mercer Investments (Australia) Limited
• Year: 2007
• ISBN:
  9781925921922
• Edition: 1
• Topic: Business
• Subtopic: Trading
• Index: Business

The volume of institutional money competing for a home is rapidly escalating, leaving in its wake a vast array of new financial instruments.

With the growing number of investment industry players and the progressively more sophisticated technology they are applying, finding exceptional skill is becoming increasingly difficult. Only those able to adapt faster than their competitors will maintain their leading edge in this dynamic environment.

Against this backdrop of investment evolution, institutional investors may ponder where new opportunities are to be found during the coming decade. Through twelve illuminating conversations with some of the world’s leading industry and academic experts, the author helps us explore the investment approaches that may be useful in generating tomorrow’s returns. He presents the reader with thought-provoking insights on gaining 20/20 investment vision.

PART I: A NEW
INVESTMENT
FRAMEWORK

1. RETHINKING THE INVESTMENT ABC

An interview with Ray Dalio on changing paradigms in investment management

“It is always from a minority acting in ways different from what the majority would prescribe that the majority in the end learns to do better.”
– Friedrich August von Hayek (1899–1992)

INTRODUCTION

Faced with diminishing expected returns on equities, retiring baby boomers (at least 76 million in the US alone), and underfunded pension plans, plan sponsors are increasingly becoming aware that while large-cap equities may take up to 60% to 70% of assets, they also account for up to 80% to 90% of the variance in plan returns. The present default option for most pension plans firmly focuses on accumulation mode, and this option had merit and support from the bullish financial markets up to the year 2000. Post 2000 however, trustees are gradually becoming aware that the more efficient developed equity markets to which the bulk of the money is allocated may not necessarily be the best asset class from which to source alpha and beta. The equity style proliferation of the 1990s may be replaced by the need for capital preservation and a focus on lower volatility products offering absolute returns.

While disentangling all these issues, corporations in G-7 countries are turning from net borrowers to net savers, freezing pension plans and increasingly transitioning to defined contribution plans, while widely debating the merits of alpha versus beta from a wider liability matching context.

Following developments in financial engineering, a number of proponents in the industry have suggested returning to the basics on portfolio construction building blocks in terms of alpha, beta and costs, in order to improve the Sharpe and information ratios, and in this interview we discuss the reengineering of alpha and beta in detail.

A paradigm shift

An existing paradigm often prevents us from seeing new ideas. Though investors have had investing tailwinds over the past 25 years, periods of rapid capital appreciation in real terms occur much less frequently than often thought, as we have described in the introduction of our book.

In the existing paradigm¹, using modern portfolio theory, asset classes are combined based on expected returns, risks and correlations. The higher expected return attributed to stocks, and the inherent home bias, carry within them a number of constraints to creating an optimal portfolio. For example, markets that have become more efficient and more richly valued at present (such as developed equity and bond markets) still comprise the bulk of most portfolios. At present for most investors, the majority of returns come from beta (generally around 95% of their risk budget), with the 5% alpha allocation being dominated by equity sources.

The new paradigm is likely to involve separation of alpha and beta, and new efforts to optimise each individually before settling on the best combination. Alpha may be used through a separate portfolio or through an overlay independent of beta, and the traditional distinction between asset classes may slowly disappear as increasingly equity and bond managers are adopting unconstrained benchmarks.

Hence, the future of investing may consist of efficient beta creators (such as indexers and ETFs) and alpha generators who will compete for the whole spectrum. In that sense, the line between traditional equity and fixed-income managers on the one hand, and hedge fund managers on the other hand, may blur in the near future as everybody starts to compete for the same opportunities.

Introducing Ray Dalio

In an era where Julian Robertson and George Soros have left the scene, Ray is one of the largest global macro money managers still standing. A baby boomer himself, Ray has been trading since the age of 12, buying and selling stocks while caddying at the local golf course. After graduating from Harvard Business School with an MBA in finance and with two years of work experience, Ray founded Bridgewater at the age of 25 in a spare bedroom in his apartment on East 64^th Street in New York, specialising in managing credit and currency exposure. Ray remains Bridgewater’s Chairman and Chief Investment Officer today, and Bridgewater has evolved into one of the largest global macro players worldwide.

The concept of alpha/beta separation was introduced by Ray over 17 years ago, but it was only after 2000, as equity markets corrected, that the idea increasingly gained traction. We talk to Ray about alpha and beta, as well as some of the deeper currents that move the markets.

Ray, can you define for our readers what you consider as “pure alpha”? What defines to you the ultimate alpha engine? Is it a lack of factor bias, or the sustainability of the uncorrelated alpha sources over time, or something else?

As you know, alpha is the return that comes from winning managers taking money from losing managers in a competition that is zero-sum. This zero-sum competition is much the same as a race, or most competitions, in that there are better and worse. So alpha being zero-sum does not mean an individual manager’s returns are zero-sum. On the contrary, as with most competitions, in the competition for alpha there are some consistent winners and losers. And, as with most competitions, the winners are better because they work harder, devote more resources to the effort and/or are more talented than their competitors.

“Pure alpha” comes solely from one’s ability to better assess what’s good to buy and sell than most other market participants. Since we are equally willing to be long or short any market or market spread (that is, any beta) based on whether it is cheap or expensive, pure alpha returns are uncorrelated with any beta. We don’t want to have betas in our return streams because they have relatively low returns relative to their risks relative to our alphas. Therefore, in addition to being uncorrelated with betas, pure alpha should have a higher return-to-risk ratio than any betas.

In my opinion the ultimate alpha engine has to produce lots of good, uncorrelated alphas. So it must be based on:

1)	Deep and logical understandings of several markets (so that one can understand what’s cheap); and
2)	A sound understanding of portfolio construction and risk management to properly bring these alphas together into a portfolio of them.

As mentioned, because I believe that “betas” and “factors” are destined to have low or nil alphas relative to their risks (because making exceptional returns can’t be as simple as doing the same thing over and over, and because everything becomes both cheap and expensive at times), my ultimate alpha engine has none of them in it. Also, because I believe that all criteria for investing (that is, good betting strategies) should have a logic that isn’t time specific, I believe that the alpha generators that make up the ultimate alpha generator should be timeless and universal. By that I mean that they should have worked over very long time horizons and in all countries’ markets.

You’re well-known for jotting down trading rules on yellow pads, which eventually became the blueprint of your trading system. Are you still doing that?

Yes, though the process happens much more efficiently than it used to. In the past, from the point of thinking of a trading rule to testing it took a long time (to gather the data and test it), so that it was the exploration of a concept which bottlenecked my progress. Now, it’s the opposite – the gathering and processing of the data is so fast that my intellectual ability to process the information is the bottleneck. The game of investing has gotten much more sophisticated since I started jotting my rules down on yellow pads!

One of the theories you suggested was the post–modern portfolio theory (PMPT).² Has the idea gained further traction within the academic or broader investment community?

The PMPT is based on my belief that the best way to construct a portfolio is to independently create optimal alpha and optimal beta mixes and then combine them. Since alphas and betas can be hedged away or taken on, and they can be engineered to be larger or smaller (synthetically), one can create very well diversified alpha and beta portfolios made up of return streams that are chosen because of their attractiveness, unbiased by whether their betas and alphas are in the same asset classes. Approaching portfolio construction this way can roughly quadruple investors’ risk-adjusted returns.

Extremely rapid progress is being made on creating the optimal alpha piece of this now. In fact, it is probably the hottest thing now going on in investing. I believe that when the alpha restructuring is largely completed, investors will turn their attention to creating the optimal beta piece with the same zeal. Then the PMPT process will be complete. Currently we are running optimal beta portfolios for about 40 clients; many more clients are exploring it and we are now seeing other managers come out with copy-cat products. So, the optimal alpha piece of PMPT is very rapidly being adapted, and the optimal beta piece is being applied in a number of cases and it is being broadly discussed. I expect that in another five years or so, the PMPT approach will be pretty standard and the traditional approach will be considered pretty archaic.

Ray, the following table shows some of the main characteristics of alpha and beta, as indicated by you.

Alpha

• Alpha sources are n...