UPS is looking for smart beta to deliver

The $28 billion United Parcel Service Inc. defined benefit plan has taken smart beta to a new level, allocating 40% of the equity portfolio to the alternatives to market-capitalization indexes.

The Atlanta-based company has some 10 different smart beta portfolios across U.S., international, emerging and frontier markets, Syed Haque, portfolio manager-global equities, said in an interview.

In total, the UPS smart beta allocation amounts to almost $5 billion of the pension fund's $12 billion in equities worldwide.

UPS executives developed the portfolios in part to avoid the concentration typical in market-cap-weighted portfolios while also seeking to lower volatility and capturing — depending on the portfolio strategy — expected return premiums for risk exposures to such factors as value, size and quality.

UPS has 49% of its smart beta allocation in the U.S. market, 37% in international markets, 10.5% in emerging markets and 3.5% in frontier markets, said Mr. Haque, who oversees the allocations.

For now at least, UPS is nearing completion of hiring managers for its smart beta allocation, which it has been developing since 2011.

Some of the smart beta indexes UPS is using were developed internally. Others came from providers such as MSCI Inc. and FTSE Group. UPS has implemented all of them into passive portfolios by the pension fund's external money managers that Mr. Haque declined to name.

UPS turned to smart betas seeking low-cost strategies to capture factors identified as having risk premiums. “Every basis point of fee is important,” Mr. Haque said.

UPS pays smart beta index fund managers from 3 basis points to 15 basis points — more than fees charged for lower-cost market-cap-weighted index funds but lower than active management costs, said Mr. Haque.

Mr. Haque also is watching for unintended exposures that smart beta might bring to both the diversification of strategies and the overall fund.

“Anytime you have smart beta portfolios, non-market-capitalization-weighted portfolios, you will induce certain tilts in the portfolio. It could be a small-cap bias, it could be certain factor exposures you (do) not intend to have ... introducing a huge tracking error,” Mr. Haque said. “You have to understand the risks you are bringing the moment you move away from cap-weighted indexes” to make sure you are being compensated.

Asset owners need to understand the place non-cap-weighted investments have in the overall portfolio, said Mr. Haque and others studying the topic.

Value, momentum and quality are among smart beta factors that have done well in the recent past, relative to a comparable cap-weighted benchmark, Mr. Haque said. Minimum variance or volatility-based smart beta portfolios haven't performed as well, Mr. Haque wrote in a follow-up e-mail.

UPS has taken the approach of diversifying smart beta strategies.

“We do switch between factors and re-evaluate factor exposure depending on market conditions,” Mr. Haque wrote.

“The overall goal of the factor portfolio is to capture different risk premias in the long run, smooth out performance in different market conditions and minimize the estimation and model risk that comes from any optimization based strategies.”

Noel Amenc, director of EDHEC Risk Institute and CEO of its ERI Scientific Beta unit, both in Nice, France, agrees.

Mr. Amenc — who has consulted with Mr. Haque — also suggests an application of smart betas, beyond a single strategy, to achieve what he calls smart beta 2.0.

“You will not achieve (relative risk management) with one index. You will achieve that with a diversification across smart betas. So you will have the benefit of (many) smart betas, invested in a well-diversified index ... (and) better risk-adjusted than a normal smart beta strategy,” Mr. Amenc said.

Eugene L. Podkaminer, senior vice president, capital markets research, Callan Associates Inc., San Francisco — who wrote a paper on smart beta Callan released in June — said smart beta investing is predicated on asset owners holding certain beliefs. For instance, “if you really believe that there is a premium for holding value stocks or high momentum stocks or low volatility stocks,” it can be captured with factor investing.

Mr. Haque said UPS sought to build smart beta passive portfolios incorporating biases toward certain factors, such as value, small cap, minimum volatility and quality, which consider the sustainability of corporate earnings.

“We do have some portfolios where we try to combine (factors) like (low) volatility with quality,” for example, Mr. Haque said. “We are getting compensated for that risk ... and avoid concentration. We don't want to have a very concentrated portfolio.”

Mr. Amenc said using diversified smart beta portfolios can overcome shortcomings. A smart beta that tries to maximize, for example, value exposure could minimize diversification.

“What I call smart beta 1.0, the first generation of smart beta indexes, are strategies that either do not explicitly control factor exposure or do not try to provide well-diversified indexes,” Mr. Amenc said in a follow-up e-mail. “When you use (smart beta) indexes, these indexes don't have any objective of diversification and moreover do not use the correlation between stocks. (Smart beta alone) doesn't try to maximize diversification. It just tries to maximize your exposure to value with a selection of criteria that produced good performance in the past.”

Other smart betas “try to achieve a diversification objective. But they don't really control for factor exposure, (which) means that sometimes you are exposed to factors you aren't looking for” and that constitute an important driver of the index risk, Mr. Amenc said in the interview.

“I'm not saying smart beta is not (capable of achieving) outperformance,” Mr. Amenc said. “I'm just saying it's not a free lunch.”

The solution in general is to “try to diversify the risk of your smart betas,” Mr. Amenc said, advocating for a new generation of indexes called “multibetas.”

“This new generation of smart beta indexes combine risk control and ... the choice of exposure to the right factors with the benefits of diversifying away the specific risk” investors don't want, Mr. Amenc said.

“The solution is to diversify across strategies,” Mr. Amenc said in the interview. “We don't trust in genius. We don't trust in somebody saying "my strategy is the best in the world.'” Instead “diversify across different risk exposures.”

“This is the future of smart betas,” he said the interview. “You will achieve ... the benefit of smart betas, invested in a well-diversified index, better risk-adjusted (performance). ... Because you are exposed to different betas you have a clear limitation of tracking error” and can expectresults much better than a single smart beta strategy,” Mr. Amenc said.

Callan's Mr. Podkaminer believes smart beta might fit well with some investors. But he said one size doesn't fit all and challenges some depictions of smart beta.

“There is a lot of hype around smart betas,” Mr. Podkaminer said. “It promises to do all these wonderful things, but when you dig deeper into the construction methodology and the empirical results, you find there is a lot of overpromising.”