Who are the true pioneers of quantitative finance? Even for us here at FT Alphaville, it’s a difficult question to answer definitively. But it’s still a fun one to ponder.
Practitioners like John McQuown, Ed Thorpe and William Fouse, as well as theorists such as Harry Markowitz, William Sharpe and Gene Fama, could all be considered canonical quant godfathers. You could even plausibly argue that Benjamin Graham was the first of both worlds. Jim Simons was certainly the most successful, in terms of how much sheer moolah he managed to make from a scientific, systematic approach to investing.
But if you talk to quants of a certain generation, there’s one lesser-known name that still occasionally gets whispered about with near-reverence: Barr Rosenberg. It’s with sadness that Alphaville learned that Rosenberg passed away last month, aged 82.
Quantitative finance is an extremely broad field. Rosenberg was specifically an early mover in what is today often called “factor investing” – the exploitation of weirdly persistent market anomalies – through his founding of Rosenberg Institutional Equity Management in 1985.
However, Rosenberg’s most meaningful legacy was years earlier inventing a more systematic framework for investment risk. In an investment industry that is often loath to change, he proved its usefulness and popularised it through his consultancy Barra. Almost all quants today stand on his shoulders. As one admirer once wrote in the FT: “Many of us use his ideas without even knowing they are from him.”
Rosenberg himself often downplayed his contributions, insisting to the financial historian Peter Bernstein that “I’ve made no major contributions to the literature”. His career ended in scandal, when RIEM had to settle SEC charges that it had covered up one of its investment algorithms going bad.
Rosenberg was fined, banned from the industry for life, and instead became a fulltime yogi and spent his last years teaching Tibetan in California. For the shy, cerebral and gentle Buddhist, the downfall was painful. “It was a difficult time,” recalls Sylvia Gretchen, an old family friend. “Very difficult.”
Nonetheless, the embarrassing end to his financial career doesn’t tarnish his legacy. Barra was later acquired by MSCI, through which Rosenberg’s approach to identifying, categorising and managing risks still underpins swaths of finance. As MSCI’s chair and chief executive Henry Fernandez told Alphaville:
Through his creation of Barra, Barr Rosenberg leaves behind a lasting legacy in the world of quantitative investing. Few have advanced systematic and factor investing more than he did through Barra. His work established a language for asset owners and managers to communicate, setting standards for risk management and portfolio construction. He demonstrated how managing risk and taking calculated risks can uncover opportunities throughout the portfolio construction process. He was truly a pioneer.
The guru
Today, Barr Rosenberg is relatively unknown. But back in he 1970s and 1980s – when people first started to use computers and data processing to solve financial problems — he was an unlikely rock star.
In May 1978, the magazine Institutional Investor put Rosenberg on the cover, depicting him as a cerebral, pink-robed giant with flowers in his hair and posing in the lotus position. Around him tiny men in suits prostrated themselves in admiration.
Inside, the magazine marvelled at “the sheer megavoltage” of Rosenberg’s mind, and described a typical scene from one of the conferences his company Barra arranged for its multiplying clients at Pebble Beach, California.
As Rosenberg speaks, a hush typically falls over the audience. In the manner of sinners, heads are slightly bowed. Eyes are moist and a bit glassy. One can almost hear murmurs of “Amen, Brother” and “Praise the Lord . . . “
Being plugged into Barr is now considered by many managers a sine qua non of a winning marketing pitch. (Nobody these days would ever ask, “Barr who?” Like Cher, he is a one-name celebrity.)
There was little in Rosenberg’s background that would indicate that he would become a one-name financial celebrity. He was born in 1942 to a father who taught Shakespeare at Berkeley and a mother who wrote poetry. He grew up as a precocious only child, enrolling in Berkeley at just 16. But in his junior year he chanced upon economics, and was entranced.
He proved a gifted thinker and mathematician – handy traits for a budding economist. As a senior, Rosenberg was allowed to attend the graduate school’s statistics course, taught by future Nobel laureate Gerard Debreu. He went on to get a masters in mathematical economics at the London School of Economics, and a doctorate from Harvard.
But finance fascinated more than macroeconomics. Uncertainty, risk and how to measure and manage it became Rosenberg’s particular focus. As he later told the New York Times:
I’ve always been interested in rational behavior in the face of risk because of illness in my family, which made me aware of the underlying uncertainties of life, and because of the history of both my grandfathers. One was a missionary, the other was an entrepreneur who went bankrupt several times. It always seemed natural to me to be concerned with the possibility of change and of rational response to very uncertain environments.
After Harvard, he joined faculty of his first alma mater, often called “Berzerkeley” for all the countercultural weirdos it attracted. Culturally, Rosenberg fitted in well, given his Buddhism and interest in original-language Tibetan literature. At work he mostly spent his time building esoteric databases of financial data and deploying the econometric methods he had learned on them, thanks to a prestigious grant from the National Science Foundation.
But when Rosenberg’s efforts to turn an old tugboat into a permanent floating home for him and his wife June proved diabolically expensive, he realised he had to start making a bit more money. So he set himself up as a consultant to the finance industry, with Dean Witter as his first client. It proved the first of many.
Barr’s Bionic Betas
It was a heady time for financial economics. When Harry Markowitz had in 1952 first articulated his Modern Portfolio Theory – a method of how to design an optimal portfolio of stocks – only the US military had computers powerful enough to actually run his equations. But by the 1960s they were becoming more available, leading to a wave of theory testing and experimentation.
Moreover, in 1964 the University of Chicago’s Center for Research in Security Prices had published a fabulous data set on US stock market returns going back to 1926. The data was collected on a magnetic tape that if unreeled would have stretched for more than three miles, and provided the raw fuel for an explosion of research.
The result was the birth of Gene Fama’s Efficient Markets Hypothesis, Fischer Black, Myron Scholes and Robert Merton’s options pricing model, and the “capital asset pricing model” developed by the likes of Bill Sharpe, Jack Treynor and John Lintner. By the early 1970s there was a surge in interest in the implications.
Rosenberg built on all this work, and helped transform it from academic theory and into practical applications. CAPM posited a single market factor that drove returns – denoted as the Greek letter beta in Sharpe’s equations – but Rosenberg suggested that there were actually dozens of different factors that in combination drove stock market movements.
For example, JPMorgan’s share price might ebb and flow with the broader market, but it also moves alongside other banks thanks to sector-specific phenomena like interest rates and regulatory changes. Then there are idiosyncratic factors, such as the skill of the CEO.
By 1974, Rosenberg’s consulting was taking off to such an extent that he founded a business to do it full-time, which he named Barr Rosenberg Associates, or Barra. The 60 or so factors that he had isolated became known as “Barr’s Bionic Betas”, and started proliferating in the investment industry.
Rosenberg’s models were a big leap from the first ones developed by the likes of Markowitz and Sharpe, as Ronald Kahn – head of quantitative research at BlackRock and a former Barra employee – wrote in a book on investment management.
They accurately forecasted risk, an improvement over the one-factor market model. They provided a coherent risk framework for investing, neatly organizing the various places that investors could try to outperform: betting on industries or factors, focusing on idiosyncratic (individual security) returns, or some blend of the two. The factor models simplified the calculations required by Markowitz optimization. Overall, Rosenberg accelerated the adoption of modern portfolio theory and, especially, placed risk at the center of investing.
Peter Bernstein’s book Capital Ideas details how Rosenberg was always focused on “applicability”. They should be useful, not just academically interesting. This practical focus helped attract the likes of the American National Bank in Chicago, Prudential, Wells Fargo, Travelers and the College Retirement Equities Fund as Barra clients by the late 1970s.
This despite what was a phenomenally complicated approach for the time – Barra’s computer programs involved about 100,000 separate instructions, and the manual Rosenberg wrote to help clients was over 1,000 pages long. Barra soon began adding other aspects to its business, such as computer programs that allowed institutional investors to analyse the funds they invested to a far better degree than had ever been possible.
Some clients therefore dubbed Rosenberg “the accountant of risk”. Here’s how Rosenberg later described Barra’s business himself:
Barra provided four core services, all of which were somewhat unprecedented but not surprising in the sense that there was a recognized need for these services.
First, providing an investment risk model that predicted the variances of individual stock returns and covariances between the returns of all individual stocks in formats that allowed calculation of risk exposure for any portfolio of stocks, which might be an existing portfolio or a hypothetical portfolio.
Second, characterizing an existing or hypothetical portfolio in terms of these risk exposures, calculating a portfolio’s overall risk exposure and attributing it to exposures to individual stocks and other factors of return.
Third, providing a portfolio optimization system that could calculate the best solution for updating a portfolio by making buys and sells so as to optimize the trade-off between reward and risk.
Fourth, attributing portfolio investment performance to the operative factors maintained in the model and analyzing historical performance to appraise cumulative results.
Rosenberg’s ruin
Rosenberg’s fall was sudden and shocking to everyone who knew him.
In 1985 Rosenberg co-authored a paper detailing what he called “persuasive evidence of market inefficiency”, which detailed the returns of systematically investing in cheap stocks (and momentum trading) almost a decade before Fama and Ken French’s own landmark paper on the subject.
He then left his own consulting firm to found Rosenberg Institutional Equity Management (Barra was later sold to MSCI in 2004 for $846mn), to put all this into practice. It was a tremendous success, accumulating $9bn of assets in just its first five years, according to Bernstein. In 1998 the insurance group AXA bought a controlling stake for $125mn, and by 2007 AXA Rosenberg managed over $135bn.
However, the financial crisis first shredded the performance of many of its models, leading assets to gush out. Then someone at AXA Rosenberg in June 2009 noticed a problem with one of the investment manager’s risk management models. Rather than fix the mistake immediately, Rosenberg said that it should be corrected as part of the release of an updated risk model scheduled for September 2009, and failed to tell the board or the company’s CEO or CIO about it.
The SEC argued this delay caused $217mn in losses for AXA Rosenberg’s clients, and excoriated Barr Rosenberg in the subsequent enforcement case:
“Rosenberg chose concealment over candor, and in doing so selfishly served his interests over those of his clients,” said Robert Khuzami, Director of the SEC’s Division of Enforcement.
Bruce Karpati, Co-Chief of the Asset Management Unit in the SEC’s Division of Enforcement, added, “Investors in quant funds trust their advisers to develop, maintain and operate the quant models that drive a fund’s performance. Rosenberg betrayed investors when he failed to disclose the material coding error.”
The fizzling performance and the subsequent scandal deflated AXA Rosenberg’s assets to just $20bn. Rosenberg resigned humiliated, and the rump of his company became a wholly owned subsidiary of AXA (which finally killed the Rosenberg brand in 2022).
Some admirers felt Rosenberg had been treated harshly by authorities keen on simply getting scalps in the wake of the financial crisis. The $217mn estimated loss from the coding error calculated by Cornerstone Research was over a two-year period and affected about 600 client portfolios, coming in at about 22 basis points on average annually. It also found that over half of AXA Rosenberg’s clients were either unaffected or actually gained from the additional risk that the faulty model led to.
Nonetheless, having been banned from ever working in finance again, Rosenberg retired to spend the rest of his life at the Nyingma Institute, a Tibetan centre in Berkeley. Even after retiring as co-dean Rosenberg kept teaching there until a month before he passed away in early February.
His wife June passed away a few years ago from cancer, and the Rosenbergs had no children. But Sylvia Gretchen, his co-dean of many years at the Nyingma Institute, says that he leaves a huge legacy among his extensive family of friends there, and will be buried on the grounds. She told Alphaville:
He had a towering intellect, and an ability to work with ideas, put them together in new ways and apply them to practical problems. The way he solved problems was really interesting.
But most of all he was absolutely delightful. He was a very kind person, who really cared about other people . . . He lived a good life, he really did. He did the best he could with the gifts he had, and he was loved. And that’s all he ever wanted out of life.
https://www.ft.com/content/66a7c514-6a74-4fe8-aba5-03dfd2146cf1
