If financial engineering is the new religion of institutional investing, then Raymond T. Dalio is the high priest. The president and chief investment officer of Bridgewater Associates Inc., Westport, Conn., has been preaching the advantages of separating alpha from beta for years. And now it's paying off: Nearly 70% of the firm's $31 billion in new mandates since the beginning of 2002 have come from overlay strategies — primarily Pure Alpha, the firm's hedge fund, and its currency overlay product. Bridgewater's All Weather strategy, which is geared to beat a standard 60% stock/40% bond asset mix, also has been a winner. Last year, assets under management jumped 46.7% to $54.9 billion. Since Jan. 1, growth has exploded another 37%, to $75 billion. Mr. Dalio spoke to Chief of Bureaus Joel Chernoff about financial engineering and the future of the money management industry.
Radical shift
A When I was growing up, I used to caddy. It was in the '60s, and everybody was talking about the stock market. … I got addicted to markets because the first share I bought was a company by the name of Northeast Airlines. My investment criteria were that it was the only stock that I had heard of that sold for less than $5 a share. It turns out that it was a company that was about to go broke, but (when) somebody acquired it, it tripled.
1966 was the first year that the yield curve became inverted, and we had a bear market, and I learned to trade from the short side as well as the long side. Then I also found that the markets were overwhelmingly dominated by macro factors, particularly related to the credit markets. In other words, which stock you bought didn't influence your performance as much as whether you should buy or short stocks in the first place.
The summer of 1972, when I was at Harvard Business School, I wanted to learn about commodities, so I went to the director of commodities at Merrill Lynch and I begged for a summer job.
A The two biggest things that are wrong are that the portfolios are not diversified, and the likely expected returns are below the requirements.
They're not diversified because the average pension fund's returns are 95% correlated with equities. Despite all of the things they do, the complexities that they add trying to (create) diversification, they're still 95% correlated with equities. They have a watered-down equities portfolio, and that's because equities make up the largest piece — nearly two-thirds of the pension fund.
A The correlations are unstable, but they've averaged about 0.4. Recently, they've been negative. But they don't have any volatility that could offset the volatility of stocks. So a 50-50 bond-stock portfolio will be dominated by equity volatility. Also, bonds have projected returns that are lower than equities. So when one diversifies into bonds, one gets a lower return and doesn't successfully diversify the equities unless they put a whole lot in bonds.
Pension funds … need 10 or 15 or 20 uncorrelated return streams that have average returns in the vicinity of 8% to 10%. Once they describe that as their mission, it leads them down a different path.
What you're seeing now in a lot of trends in the industry is an extension of those two problems: the absence of diversification and the low returns … Fifteen years ago, we started separating alpha and beta. Nine years ago, we started the "All-Weather" portfolio mix. And we would show clients these strategies, and they weren't very interested because the traditional approach paid them well. Up until the year 2000, there was no reason to change the strategies.
But everything has really changed since 2000. So now you're seeing a radical shift in how money is being invested. There almost was like a light switch (turning on), where the separation of alpha and beta, using portable alpha, using hedge funds, using leverage to balance the asset allocation mix better.
A Three years ago, 85% of our mandates were not structured by taking pure alpha — using an overlay on a benchmark. They were structured the traditional way, even though we gave people the choice. We made clear the preferable choice was this alpha overlay. Now, 70% on average are structured the other way. Clients can pick any benchmark, any beta — any asset class that they want, and we will passively replicate that. And then we will overlay it with our pure alpha, which is our best alpha mix. And we will calibrate it to their desired tracking error. It's like a Chinese menu: Column A is the benchmark, Column B is the alpha. And then (we) say, how spicy do you want the alpha?
A Let's say I'll give you an S&P 500 beta, and (you) put pure alpha — your best mix — on top of the S&P 500, and do that at a 3% tracking error. Or it could be 12% risk or 1% risk.
In some cases, we're given an allocation at the total fund (level). Say we'll have 25 basis points' risk at the total fund. Risk budgeting is very much a part of this. Their equity piece is maybe 60% of their portfolio and maybe has an 18% tracking error.
So that means that it has almost a 12% total risk, in risk budgeting terms, of the total portfolio. Wow, they have 1,200 basis points in volatility coming from equities. I can give 25 basis points of volatility or risk to Bridgewater on a total fund level.
A You can take an inflation-indexed bond and leverage it to have the same expected return as equities. And now if I put 50-50 of my money in each of those two investments, I have a much better portfolio.
But if I don't alter it, through the use of leverage, I'm caught in a dilemma when I invest in inflation-indexed bonds: I have to lower my expected return, and I don't have an effective balance to my equity volatility.
While we could debate just what is the expected value of asset classes, it's almost immaterial to the question because, just by leveraging asset classes, or by leveraging alphas, I'm able to get them close to the same size.
A We literally have well over 100 sources of alpha. Alpha requires you to have a deep understanding of market inefficiencies. We have to research and understand more about markets than anybody in our business. For 29 years, we've been particularly focused on the credit and currency markets. Competing for alpha is more challenging than competing in the Olympics. What we do with those alphas is that we have a lot of different ones and we create a portfolio of those alphas.
You want to have enough diversification so that no significant portion of your bets can be wrong. Our Pure Alpha (strategy) has had for 15 years an information ratio that's been 1-to-3. The individual components of that have information ratios about 0-to-35 on average. But because they're diversified, they produce 1-to-3.
