Pimco’s 125 investment specialists meet quarterly to vote on which fixed-income risk factors to hedge and exploit. The firm’s risk-centric decision-making has allowed it to consistently beat its benchmarks.

Pacific Investment Management Company (Pimco), the Newport Beach, California-based fixed-income asset manager, has consistently beaten its benchmarks. Pimco’s success is based on a complex risk analysis that allows it to sift thousands of risk factors in order to locate and hedge dangerous exposures, and identify attractive investment opportunities.

Its $50 billion Total Return Fund has returned an annual average of 150 basis points over its benchmark, the Lehman Brothers Aggregate Bond Index, for the past 10 years, and has had an historical tracking error of 102bp a year since inception.

Every quarter, Pimco investment managers gather for days in conference rooms to debate which portfolio risk factor weightings the firm should use and which relative value trades it should pursue in the following three months. The debates culminate in a series of votes, weighted according to seniority and senior management’s view of the voter’s knowledge of a given investment issue. Pimco’s senior investment managers can add or discard risk factors to guide the process. “The real trick is how to identify five, seven, 10 or 12 factors that capture 99.9% of the risk,” says Vineer Bhansali, head of the portfolio analytics group, where Pimco’s risk management function is centred.

Over the past year and a half, Bhansali has developed a mathematically rigorous method for combining Pimco’s investment specialists’ votes on factors with historical risk factor correlations – in essence hybridising them. The weaker the confidence of Pimco’s investment team in forecasting a given correlation, the more weight is assigned to historical correlations, and vice versa. In the past, researchers seeking to generate useful risk factor co-variance matrices have baulked at the enormity of forecasting as many as 5,000 risk factor correlations. Pimco’s approach reduces the strain on analysts by concentrating portfolio managers on just the most important variables. Among the 14 risk factors currently in the spotlight at Pimco, the most important relationship now is between yield curve shape and spreads.

The results of the hybridisation are a major part of the firm’s strategy for a given three-month period. The strategy is input to a desktop risk management system Bhansali and his team developed in-house, called Blotter. With Blotter, 60 or so security-level risk analytic tools and several aggregate measurement techniques for liquidity and credit risk are overlaid on the firm’s strategy. On a daily basis, Pimco’s portfolio specialists use Blotter to identify cheap and rich securities. This has played a big part in Pimco’s consistent success.

From a client perspective, according to Bhansali, the hybridising approach offers greater ease in assessing the impact of stress events on different portfolios’ risk factor exposures. “In periods of stress and liquidity, certain correlations might go to extreme values, and historical analysis will never get there at all,” says Bhansali. “In a way our methodology expands the possible universe of outcomes beyond what has happened in the past, and lets you imagine ‘bad’ scenarios in a consistent manner.”

Pimco has been known for making big bets, so its risk management emphasis sometimes comes as a surprise to clients. “I came away with something different from my initial perception a couple of years ago when we hired them, which is they tend to be big risk-takers,” says Dwight Kadar, investment management director of a $1.2 billion pension fund for Cooper Industries, a manufacturing company based in Houston, Texas.

In 2000, lead portfolio manager and Pimco founder Bill Gross made headlines with the $8 billion position he took in long-dated US Treasuries, Treasury futures and zero-coupon bonds before the rest of the market had fully realised the implications of the US government’s debt repurchase plan on the prices of long-dated Treasuries. Large bets, however, are not representative of Pimco’s approach. “Given that there’s a relatively asymmetric risk in bonds against you, you’re much better off avoiding the blow-ups than you are trying to find the winners,” says Shannon Bass, investment-grade corporate bond specialist at Pimco. Cooper Industries’ Kadar recently spent several days at a Pimco seminar in its Newport Beach, California headquarters. “Their risk control is very conservative as a firm,” he says. “They use a lot of exotic instruments for risk control purposes.”

Other clients gush over how secure they feel with the depth of Pimco’s risk systems, but for them the story really lies in Pimco’s returns. “The performance has been outstanding relative to the benchmark,” says John Krimmel, chief investment officer of the $10.5 billion Illinois State Universities Retirement System, of which Pimco manages $2.2 billion. “They were around 160 basis points over the benchmark, yet their risk profile was only slightly higher than the Lehman aggregate.”

The $12.4 billion University of Texas Investment Management Company (Utimco) runs its own fixed-income portfolio except for a roughly $915 million allocation to Pimco. Russ Kampke, senior portfolio manager for Utimco, says: “With the market moving around quite a bit in the last year, obviously their ability to make their bets and know where they are versus the market has helped them, because they’ve got good numbers.”

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here