Fractal Strategic US Treasury Fund: Fractal Advisors

Fractal Advisors has turned a simple idea into a winning strategy. Hedge Funds Review speaks with managing director Jorge Zighelboim about Fractal’s unique approach to investing in US Treasuries.

The best ideas are often the simplest, believes Jorge Zighelboim, managing director of Fractal Advisors. He thinks this maxim applies to the Fractal Strategic US Treasury Fund (STF), one of the most uncomplicated products in the hedge fund universe.

“I can explain what we do in five minutes. There is no mystery to it,” Zighelboim says.

The Fractal STF invests exclusively in one asset class: US Treasuries. The fund is structured as a multi-manager platform. However, around 90% of its assets are allocated to a pair of quantitative long/short strategies, one based on a proprietary model and the other managed by a third party.

The aim of the fund is to generate absolute returns in rising and falling interest rate environments. This is primarily achieved through long and short positions on US Treasury futures. Individual positions are based on signals generated by quantitative models which aim to predict future interest rate movements.

Fractal’s core long/short model aggregates macro-economic data from the US on a monthly basis, from the first Friday of the month when the unemployment numbers come out until the next set of data is released a month later. Based on this data, the model aims to predict whether interest rates will move up or down.

Each month the fund takes either a long or short position in US Treasury futures based on these signals, or goes to cash if the model is inconclusive. It all adds up to just 12 positions each year.

Zighelboim and Alain Bibliowicz, his fellow managing director at Fractal Advisors, ran a similar strategy at UBS on the sell-side where they worked alongside the bank’s quant team. At the bank the model was used to make long-only investments in US Treasuries.

In 2005 Zighelboim and Bibliowicz left the bank and began developing a new model which could generate both long and short signals using a similar methodology to the one used at the bank.

“We realised that a long/short approach could increase returns by up to 5%–6% a year,” says Zighelboim. Later that year they launched Fractal to trade the model.

Fractal allocates around 10%–30% of its capital to its proprietary long/short model. A further 40%–70% is allocated to another model operated by a third party.

“The two long/short models are very similar, but they sometimes produce contradictory signals. Having an allocation to both models provides some diversification to the portfolio and helps to reduce volatility,” says Zighelboim.

In any given month 85%–90% of the fund’s capital is split between these long/short models. The remainder of the portfolio is allocated to spread and momentum strategies, “which add some alpha to the fund”, says Zighelboim.

“The standard deviation of the spread and momentum strategies is much higher than the core long/short strategies. That is why they have a smaller allocation. If we had more exposure to these models it would significantly change the risk/return profile of the fund,” Zighelboim explains.

Fractal uses a proprietary asset allocation model to determine the monthly optimal allocation to each model. The different strategies allow Fractal to exploit the volatility resulting from rising or falling interest rates, short-term inefficiencies and changes in the slope of the US Treasury yield curve.

The appeal of the fund lies in the accuracy of the models and their ability to forecast correctly future interest rate changes. The portfolio has historically had a hit rate of almost 72% in terms of positive months. The models have a success rate of almost 70% when it comes to predicting correctly the future direction of interest rates.

Zighelboim stands behind the reliability and consistency of the models. “The strategy is very repeatable. It is one of the reasons we chose the name Fractal. In nature, fractals are infinitely self-repeating patterns, like snowflakes or the leaves of a tree. One leaf looks like the next. Our aim as a hedge fund is to produce the same results time after time, to never deviate or become different from what we are,” he says.

Zighelboim sees the Fractal STF as an alternative to long holdings of US Treasuries. “Treasuries are fantastic to have in a portfolio along with other fixed-income positions when rates are going down, but terrible when rates go back up,” he notes.

The result is a costly cycle whereby investors build up huge exposure to US Treasuries during low interest rate environments, only to flee when rates start to rise. “That is what we try to avoid,” says Zighelboim. “What we have done is create a product that allows investors to stay in Treasuries irrespective of whether rates are going up or down.”

Rather than relying on rising prices to generate returns, Fractal aims to capture the volatility of US Treasuries. “The more volatile the market, the better our returns should be,” Zighelboim remarks.

He thinks investors typically underestimate the volatility of US Treasuries. As an example, he points to the CBOE 10-Year Treasury Yield Index, which started 2009 at 2.4 and rose to almost 4 in June, only to drop to 3.2 in December.

“US Treasuries are extremely volatile, but people think of them as safe investments because the trading range is usually between three and four. But if you look at it in percentage terms, that is a 33% gain on the way up and 25% on the way down,” he says.

“Looking at the second half of 2009, the trading range has been between 3.2 and 3.8. Those are pretty big moves,” adds Zighelboim.

Fractal has thrived in the recent market volatility. It generated positive returns of 6.74% in 2008 and was up 5.71% to the end of November 2009.

The fund has generated annualised returns of 4.25% since inception in October 2005, with a standard deviation of 6.03%. The Fractal STF has a -0.21 correlation to the S&P 500 and correlation of -0.15 to the 10-year US Treasury bond.

Sun and rain
“I describe it as an all-weather fund,” says Zighelboim. “We make money regardless of whether Treasuries are going up or down.”

On the downside the fund is mainly exposed to interest rate and extreme event risks. Credit and liquidity risk are non-existent in the sovereign debt market. The volatility in interest rates allows the fund to make returns. It can also lead to losses if the portfolio is positioned on the wrong side of interest rate moves.

Zighelboim says Fractal tries to minimise these risks by using multiple, uncorrelated models and through strict stop losses. These guard against extreme event risk. It also runs an optimisation model focusing on minimising drawdowns and maximising the returns on positive months.

A key concern is ‘model risk’. This is essentially the likelihood that Fractal’s models lose their ability to forecast correctly interest rate changes. Zighelboim says much of what Fractal does on the risk management side is geared towards monitoring and refining the accuracy of the model.

He is quick to stress that Fractal’s investment approach “is not a black box process”. The investment team monitors the models on a daily basis and tests their correlation to interest rate changes. “It is a continuous research process. We are always researching new variables which can be substituted into the model if existing ones lose their power,” he says.

Nevertheless, it is rare for Fractal to make changes to its model, he adds. “It happens once every 18–24 months.” But he sees continuous research and a willingness to make changes to the model when required as crucial to Fractal’s long-term success.

“You have to be prepared to change the variables in the model otherwise it will become outdated and eventually fail. To me that is the definition of a black box: a model that remains static until it becomes ineffective. Our aim is to ensure our model never loses its correlation to interest rates. That can only be achieved through continuous research and development,” says Zighelboim.

Concerns about inflation and the future direction of interest rates have led to increased interest in Fractal’s strategy among institutional investors. “Interest rates have hit rock bottom. They can only go up from here,” he notes. This, combined with the prospect of higher inflation, is creating a real headache for investors with significant exposure to fixed income.

“In this type of environment, investors want to exit Treasuries, but where can they go? Other asset classes have their own risks. We think Fractal offers a good solution. They can remain in Treasuries without being exposed to the risk of drawdowns if rates rise,” says Zighelboim.

The fund is steadily gaining traction among institutional investors, he says. A large institutional client recently hired Fractal to replace its allocation to the bond giant PIMCO.

“What they liked about Fractal was that our returns and volatility were in the right range through 2008 and 2009, while PIMCO has been all over the place,” says Zighelboim.

Fractal launched in October 2005 and by the end of 2006 had gathered over $67 million in assets under management, mostly from investors who had known Zighelboim and Bibliowicz during their time at the bank, as well as a seed investor from Venezuela. Fractal currently manages $164.9 million, with half of those assets coming from institutional investors.

“We attracted fresh capital into the fund through 2008 and 2009 when most hedge funds were suffering redemptions. That was very rewarding. We have opened a number of important new accounts since June 2009,” says Zighelboim.

The success of the fund is partly due to its uniqueness. “There is no one else out there who can do this successfully. It is hard to put together but once you have the elements in place, it is easy to manage. What we do is very plain vanilla,” he says.

One of the challenges Zighelboim faces is convincing potential investors to view Fractal as an alternative to long Treasury bond exposure, rather than as part of their allocation to traditional hedge funds. While the fund aims to generate absolute returns, Fractal has none of the complexities or risks typically associated with hedge funds, he argues.

“Hedge funds use all kinds of complex tricks to generate alpha. Fractal is very simple. We take a single, liquid instrument and place a trade monthly. There is no financial leverage in the portfolio. Investors have daily liquidity and transparency. There are no gates or sidepockets,” he says.

Fractal builds its portfolios using US Treasury futures. This approach has two main advantages, says Zighelboim. First, it lowers the expense ratio of the fund as the transaction cost of trading futures is much lower than purchasing cash bonds. Second, futures capture the implications of interest rate changes more accurately than cash bonds due to the extra liquidity in the market.

The use of futures also allows Fractal to offer the strategy as a portfolio overlay, an option that is proving to be popular among institutional investors. The overlay is structured as a separately managed account in the name of the investor and requires about 10% initial funding.

Fractal is required to post only a 6% margin for the futures contracts it purchases, with the remainder held to cover margin variation. “Adding Fractal in the form of an overlay can transform the risk/return profile of an institutional portfolio while leaving the underlying asset allocations largely intact,” says Zighelboim.

For example, an equities portfolio benchmarked to the S&P 500 would have generated annualised returns of 0.34% with a standard deviation of 15.99% between January 1999 and September 2009. Adding a 10% Fractal overlay to the portfolio would have increased annual returns to 5.68% and reduced standard deviation to 11.43% in the same period.

“This is a very cost-effective way for institutional investors to diversify their portfolio. Investors have full transparency and the level of exposure can be tailored to suit the needs of the investor,” says Zighelboim.

The overlay has been available to investors since summer 2009. Zighelboim says it has met with an overwhelmingly positive response from investors. “I think most of our growth over the next couple of years will come in the form of portfolio overlays to institutions,” he says.

FUND FACTS: FRACTAL STRATEGIC US TREASURY FUND

Full name of fund: Strategic US Treasury Fund
Name of investment/management company: Synchronicity Investments
Contact: Jennifer Grantham ([e-mail])
Launch date: October 2005
Assets under management: $164.9 million
Annualised return: 4.25%
Annualised volatility:     6.03%
Strategy: long/short fixed income
Administrator: TMF Group
Auditor: KPMG
Futures commission merchant: MF Global
Domicile: British Virgin Islands
Management fee: 2% charged monthly
Performance fee: 10% charged quarterly
Minimum investment:    $100,000
Lock-in/up: none
Redemption period: monthly with 30 days’ notice