Intech bases its investment process on the formula above. [Photo: Scott Wiseman]

Walking through the penthouse offices of Intech Investment Management in West Palm Beach, two things strike the visitor. One, the offices convey an astonishing blend of science and sensibility. Paneled, curvilinear halls — there's hardly a right-angle in sight — and unusually shaped decor are meant to evoke mathematical concepts. Stunning works of art, some melding equations into their subjects, adorn sprawling conference rooms and offices.

Two, the offices feel empty. And in fact, 64 people work in space created for 125 barely three years ago, when Intech had $69.7 billion in assets under management and seemed on its way to much, much more.

But that was nearly $30 billion ago.

√While its logo is visible on CityPlace's signature building in downtown West Palm Beach, Intech maintains a very low profile. "I wasn't sure I even wanted our name on the building," says CEO Robert Garvy.

Intech's up-and-down journey begins with a Princeton University mathematician, E. Robert Fernholz. He is an expert in stochastic calculus, a branch of math used to predict the behavior of systems or processes in which there's a lot of random variation — the diffusion of particles in space, for example. For a half-century, economists and analysts also have used it to study the movement of interest rates and stock prices. In today's market, stochastic calculus is used to model everything from option pricing to derivatives, hedging and automatic trading.

In 1982, Fernholz published a paper presenting insights about stock market equilibrium that he derived using stochastic calculus. Five years later, he left Princeton and used his theorem as the basis for Intech, an investment firm he founded under Prudential Insurance.

Robert Garvy became a partner in the business in 1991 and moved the company to Florida. Intech capitalizes on the volatility of stocks within a particular group such as the S&P by reweighting the companies in the index, selling the ones outrunning the index's return line and buying the underperformers. [Photo: Intech]

Fernholz's novel approach attracted the attention of an Atlanta-based investment consultant, Robert Garvy. "I was fascinated by the intellectual power of stochastic portfolio theory," says Garvy, who in 1991 became Fernholz's partner.

Fernholz and his research team stayed in New Jersey. But Garvy was a Floridian. A 1960 graduate of Melbourne High School who had lived in Florida since he was 11, "I first learned how to sail in a little pram" in Melbourne. He began his bachelor's at the University of Florida and finished at the University of South Florida before earning an MBA at Georgia State University.

Garvy moved Intech — then managing only one investment fund focused on companies in the S&P 500 — first to Titusville and then to Palm Beach County. The firm took awhile to bloom. It catered to institutional investors — pension funds, insurers, union funds — but even their sophisticated administrators found Intech's process alien.

In making investment decisions, Intech ignores all the so-called "fundamentals" — whether the economic outlook is conducive to higher sales at Walmart or whether a new Red Lobster menu item means more revenue for Darden. Indeed, Intech couldn't care less why a stock goes up and down. There are other such "quantitative" firms that ignore fundamentals. But Intech is the purest form of quant, caring only about a stock's movement relative to other stocks and the benchmark.

Profiting from volatility

Crudely put, Intech's approach is to buy low and sell high, but what to buy and sell depends on Fernholz's bedrock theorem. Intech capitalizes on the volatility of stocks within a particular benchmark such as the S&P 500 by reweighting the companies in the index, selling the ones outrunning the index's return line and buying the underperformers. When the underperformers cycle up to outperformers, Intech sells them. "Math is Power" is the company slogan.

Former Princeton University mathematician Robert Fernholz started Intech in 1987.

"They're trying to modestly outperform without the volatility," says Morningstar analyst Courtney Goethals Dobrow, who follows three Intech funds.

Intech's flagship fund aims to beat the annualized return for its benchmark index by 2 percentage points before fees. The difference between an 8% annualized return and a 10% return on a $100-million investment compounded over 30 years works out to be the difference between $1 billion and $1.7 billion, Garvy says. "Those dollars end up paying for people's retirements," Garvy says — the difference between "filet mignon and cat food" for retirees.

Denver-based mutual fund giant Janus Capital bought out Prudential's majority stake in 2002, and the filet mignon argument gained traction with retirement funds and others. But the company's eagerness for client business got it in trouble in 2009. That year, without admitting or denying fault, it paid a $300,000 penalty after the Securities and Exchange Commission faulted it for failing to properly disclose to clients from 2003 to 2006 that it was casting pro-labor votes on their behalf at companies' annual meetings. The SEC said Intech wanted to curry favor with potential and existing union clients.

Meanwhile, Intech landed the business of international institutional clients, including sovereign wealth fund clients, in Japan, Australia, Europe, China and Abu Dhabi. The cash poured in, averaging $12 billion a year in net inflows from 2004 to 2006, and Intech reached $69.7 billion in assets by year-end 2007. Intech remained a bright spot even as parent Janus struggled.

√Intech generally met or exceeded its goals until a combination of market conditions and poor performance hammered its portfolios. Assets under management fell from $69.7 billion in 2007 to around $40 billion today.

Amid that prosperity, Intech planned its move into the CityPlace Office Tower, becoming the signature tenant in the signature building in the signature development in downtown West Palm Beach. It was something of a departure from the anonymity Intech cultivated, in keeping with its publicity-adverse clients.

By the time it took occupancy in November 2008 in the midst of the great market meltdown, Intech was struggling. For much of 2006 through the huge 2009 rally, the approach that had worked so well performed poorly against benchmarks. Intech went from Janus' bright spot to its dark cloud. Its strategies — funds and accounts managed for individual institutions — lost billions in 2008: $26.5 billion as the market depreciated. Investors beat for the exits, taking more with them. As of June 30 this year, Intech was down to $39.8 billion in assets.

"It's certainly been a drag in terms of flows for Janus overall for quite a number of quarters, mostly driven by some performance issues they've had. In the past, it's been a big driver of growth for Janus at a time when the legacy business for Janus has been under some pressure, but now that's reversed," says Sandler O'Neill + Partners analyst Michael Kim.

It turns out that Intech's approach worked well when the overall level of market volatility remained relatively consistent but faltered when the market shifted radically from one amplitude of volatility to another. And because Intech generally depended on capturing the movement of smaller, more volatile players in a benchmark like the S&P, it didn't do as well when investors began to pile into the largest stocks.

√Intech doesn't care whether a company has a new product or is in a declining industry — its investment decisions are driven by mathematics that predict stocks' behavior based on their value relative to other stocks and the index.

To respond to the changing market conditions, Fernholz's theorem can no more be changed than the ratio of a circle's circumference to its diameter. But fund managers can tweak how they act upon his theorem — how many years of prior volatility data inform their buy-and-sell decisions. Garvy says that over the decades, the firm has made infrequent, tiny adjustments to increase return or lower risk, but they amount to a just a handful of basis points in performance.

"There will be periods, no matter how good you are, when you will underperform," Garvy says. Intech has spent 2010 explaining to investors that it's still good rather than no longer lucky.

Comeback

Internally, Garvy stresses the importance of keeping talented investment managers satisfied and stimulated and the firm's employee turnover low. Recent results show promise. All told, 42% of Intech strategies beat their benchmarks over a one-year period as of the June 30 quarter after fees. None did so in the prior quarter. And 83% beat the market for the three-year period, compared to just 42% in the first quarter. Janus CEO Dick Weil told analysts in July that Intech's risk controls worked well, allowing it "to come back reasonably quickly with outperformance, and that's key to their success. ... We also recognize that one quarter does not make a trend. They continue to face some very challenging industry circumstances."

Co-CEO Jennifer Young will succeed Robert Garvy when he steps down next year.

Since inception, the annualized return on Intech's original strategy is up 1.13 percentage points after fees against the S&P 500, for a 9.02% annualized return. Over time, that 1.13-point edge means real money — a gain of $154.8 million after fees over the S&P 500 on a $100-million investment made at inception. The annualized return on other Intech strategies after fees ranges from a negative 2.49% to a positive 3.79% against their respective benchmarks.

As for flows, the bleeding continues but it has moderated. After net outflows totaling $9.2 billion in the prior three quarters, net outflows slowed to $1.5 billion in the June 30 quarter.

Recapturing lost business is another matter. The sophisticated people who run huge retirement funds think just like Ma and Pa Investor. After a decade of no gain in stocks, they want more money in fixed-income investments. Particularly ill news for Intech: Casey Quirk, in its annual survey of consultants for institutional investors representing $7 trillion in assets, found demand for U.S. equities way down and found no preference for quantitative firms after poor performance. Institutional investors represent 81% of Intech's business, with the remainder mutual funds.

"Institutional demand for quantitative equity strategies has been pretty subdued" as some investors question math-driven investing strategies, says Sandler O'Neill's Kim. He says investors are going in two directions on the risk curve — either seeking more risk through alternative investments or little risk with fixed-income. He expects Intech's outflows to continue this year as institutions restructure their strategies.

Co-Chief Investment Officer Adrian Banner will become the sole CIO when Robert Fernholz retires in 2011.

Garvy and Fernholz, nearing the end of their careers, have opened ownership of their share of the firm to employees. Garvy, now 67, has a co-CEO, Jennifer Young, who will be his successor when his contract runs out next year. Fernholz, now 69, will relinquish his chief investment officer role in 2011 to co-CIO Adrian Banner, who works out of Princeton with the rest of the research team. Garvy expects to remain board chairman and Fernholz investment committee chair.

As Intech again outperforms, Garvy expects business to build. Based on assets under management, it remains the largest player in its specialty — active-managed U.S. institutional equity quant, according to eVestment Alliance. It ranks seventh among U.S. actively managed institutional equity managers. It executes 1 million trades a year from West Palm Beach.

"Twenty-three years we've been beating the stock market. This is something managers are not supposed to do with market-like risk," Garvy says. "We've had some very significant success, but I think we also have some tremendous opportunities going forward."