Mark Fiskio is founder and managing partner of Wellesley, Mass.-based Empirical Asset Management. Previously, he was senior portfolio manager at Merrill Lynch; first vice president, investments, and branch office manager at Advest; and vice president for the Investment Services Group at Donaldson, Lufkin and Jenrette.
Which single asset class is Empirical Asset Management most bullish (or bearish) about in the coming year?
Our Rules-Based Investing process has always been 100% strategic and completely eliminates human emotion from the decision-making process, hence the Rules-Based name. Empirical Asset Management uses proven sets of rules and we don’t deviate from our rules, therefore there is no subjective decision making whatsoever in our process. This makes Rules-Based Investing unique in a world of commoditized investment vehicles and programs that are typically making tactical bets that allow emotion to creep into the process.
The new trick on Wall Street is to disguise tactical bets within asset allocation models, by replacing tactical bets on individual stocks with tactical bets on asset categories. What’s the difference? There really isn’t one. This new tactical allocation business is a wolf in sheep’s clothing. It will work no better for investment firms and individuals than tactical bets on individual securities have over the years.
John Bogle from Vanguard recently conducted a study that indicated investors in ETFs are receiving 3% less return per year than the ETFs that they are investing in. Why? Fear and greed, which in our opinion are the two most destructive forces to successful investing. Investors are simply getting in and out at the wrong time based on their emotions and other human frailties and biases. They have traded bets on individual stocks for bets on asset categories. Think about this: The highest equity mutual fund redemption rate in history was from November 2008 through March of 2009. Many of those redeemers missed the 100%-plus the market has enjoyed since.
Tactical asset allocations models are all the rage on Main Street because of the terrible experience individual investors had from 2000-2002 and again in 2008 and the corresponding belief that all they were missing was a strategy to successfully time the market. Every study in history proves this notion to be wrong. Wall Street loves to sell investors what they want now, not what they need, because it makes for better immediate sales even though it is counterproductive to the long-term benefit of the client and the relationship with their adviser. Tactically betting on asset classes will prove no more successful than betting on individual stocks.
Our experience has been that opinion has little to do with investment success, so we remain agnostic at all times as to securities and asset classes. Look at the periodic chart of asset category performance [see figure] and tell me which asset class is going to perform best next year. I asked hundreds of financial advisers in 2010 to attempt to select the best performing and worst performing asset classes from the prior year, 2009. Even with hindsight not a single one of them got both right. These are the professionals. If sophisticated investment professionals with hindsight can’t do it, just try doing it with foresight. It is a loser’s game.
This does not mean that Empirical Asset Management employs only passive management. Our portfolios are both actively and passively managed, but unlike most money managers, we are actively managing the risk, not the return. We typically gain performance from the very act of controlling risk, where most managers are attempting to gain a performance advantage from the very act of assuming risk, or making bets. History proves that this method does not work well for investors.
So where does this risk-management approach lead you in terms of classes?
Although we have no opinion of which asset class or ETF is going to be the best performer this year, or any year for that matter, we do get excited about particular asset classes and ETFs. Our process is all about risk control, so when we are able to add an asset class that provides correlation benefits as well as risk/return benefits, it gets our attention.
The Alternative Investment (AI) asset class, and specifically hedge funds, has all of these advantages. Multi-Strategy Hedge Fund AI as an asset class in total has a correlation to the S&P 500 of only about 0.65 with returns that have been much better and risk that is much lower. While it would be great to add this asset class to a portfolio for these reasons, there are several obstacles to doing so, including: individual manager risk, lack of transparency, lack of liquidity, potential excessive leverage, inefficient taxation and high fees.
Regardless of these issues, because the benefits are so great, institutions, endowments and wealthy individuals tolerate these problems so that they can benefit from this asset class. It is that valuable to them. Considering all of these drawbacks, we thought it would take years to be able to add this asset class to our portfolios.
Which ETF position would Empirical Asset Management choose to capture that?
IndexIQ delivers all of the benefits of the Multi-Strategy Hedge Fund AI asset class, and without the headaches, through their IQ Hedge Multi-Strategy Tracker (QAI), a multi-strategy hedge fund replication ETF. They truly have democratized the use of the Multi-Strategy Hedge Fund AI asset class for all investors, which previously were only available to sizable institutions and very wealthy individuals. QAI captures the beta of the asset class as opposed to the alpha of individual managers. We like to say that they capture the beta of the hedge fund asset-class alpha.
Dovetailing perfectly with our process, IndexIQ uses a rules-based process to replicate the bets of hedge funds as an asset class, not the bets of any individual managers. The way they do this is by first separating all hedge funds reporting to a Non-Investable Hedge Fund Index (capturing about 85% of all operating hedge fund assets) into six buckets - Event Driven, Long Short Equity, etc. ... They then look under the hood at the performance of each bucket - again, not at the individual managers.
After analyzing the performance data they use a turbo-charged form of regression analysis to determine what investments were made in order to gather those types of returns, effectively reverse engineering the performance into likely holdings. Not to be overly simplistic, but if one of the buckets had a return in 2008 of -40%, it was most likely invested in stocks, not bonds. Of course, their process is much more granular than that.
Once they understand the positions in each bucket they aggregate them together. Maybe a long bet in an asset class in one bucket is canceled out by a short bet in the same asset class in another bucket. Once aggregated, they need to replicate those positions and they do this by investing in other ETFs. They typically hold approximately 20 ETFs in QAI.
So what has QAI tended to hold recently?
Not surprisingly the three largest ETF holdings in December 2009 were iShares iBoxx Investment Grade Corporate Bonds (LQD), iShares MSCI Emerging Markets (EEM) and iShares Barclays 1-3 Year Treasury (SHY). As of the end of January 2011, the three largest ETF holdings were Barclays Aggregate Bond (AGG), Vanguard Total Bond (BND) and iShares MSCI Emerging Markets (EEM). After performing their monthly analysis they capture approximately 65% of all hedge fund assets under management. By using this Non-Investable Hedge Fund Index they capture some funds that are open for investment, but more importantly they capture those that are closed to investors. As you can well imagine some of the best performers are closed to new investors.
Can you compare QAI to traditional hedge fund investing?
Both QAI and traditional hedge funds can deliver the benefits of the Multi-Strategy Hedge Fund AI asset class in risk/return and correlation benefits. QAI is an ETF (open to all investors) versus traditional hedge funds that are usually structured as a Limited Partnership (accredited investors only). ETFs can provide liquidity throughout the trading day while traditional hedge funds have lockups and gates, which can become a major detriment when investors are demanding liquidity. QAI does not employ leverage while most hedge funds do employ leverage. A typical hedge fund charges a 2% management fee and a 20% performance fee. Compared to approximately 1.10% all in for QAI (75 basis points of expense ratio for QAI and roughly 35 basis points for the ETFs that they hold), the ETF is a bargain.
QAI is rules-based and transparent as opposed to the secretive and idiosyncratic nature of the typical hedge fund. Currently the hedge fund world is lobbying/battling the SEC, Congress and others to remain as secretive as possible. The way the ETF is structured provides more tax efficiency. Traditional hedge funds usually generate a K-1 while QAI investors receive a 1099. Further, QAI did not throw off a capital gain in either 2009 or 2010 while hedge funds are notorious for creating short-term gains. The creation and development structure of the ETF is helpful here.
What about other hedge fund replication products? Why is QAI Empirical Asset Management’s first choice?
There are other hedge fund replication products and the field is expanding rapidly because of the benefits that can be derived. Many of these products are specific to subcategories of the hedge fund world. IndexIQ, for instance, has a Merger Arbitrage hedge fund replication product (MNA). We seek to add the broad asset class as a diversifier, so those specific hedge-fund ETFs do not fit our needs. Granulation will come later as the products and categories become more commoditized and reliable.
Most of the other replication products are not in ETF form; rather they are in open-ended mutual fund structures. Several of these funds invest in individual securities in an attempt to replicate the bets of the hedge fund managers and tend to throw off unwanted tax events. Those funds typically do not meet our discipline. Additionally, IndexIQ’s rules-based methodology makes the application of their investment process more consistent, disciplined and reliable. Their acceptance in the market place is expanding rapidly, providing the liquidity necessary for us to trade sizable blocks when necessary.
Can you tell us a bit more about your investment process, and how QAI fits into Empirical Asset Management’s portfolios?
Empirical Asset Management has five asset allocation portfolios: Conservative, Moderate Conservative, Moderate, Moderate Aggressive and Aggressive. Our portfolios combine active and passive management with precision asset allocation. QAI is our alternative component: the beta of multi-strategy hedge fund alpha.
QAI ranges from a 5% weighting in our Conservative model to a 20% weighting in our Aggressive model. The models are strategic, not tactical, and there is an active component to all of our portfolios. We use proven, long-term sets of rules designed to quantitatively identify mis-valuations in the market place. Our active core portfolio selects approximately 140 underlying securities. Once we understand the makeup of our core we surround those holdings with ETFs to gain precision asset allocation. Every 15 months we rebalance each of the portfolios, which serves to freshen up the active portion and bring risk back under control. This is all done in a tax-efficient manner, as all of the gains and losses are long term.
Rebalancing is effective because it provides the investor with two important benefits: reduced risk and better returns. Mitigating risk is accomplished by reducing overweighted categories so they have less of an impact on volatility should they decline (see large-cap growth in 2000). The second benefit is not quite as obvious, but rebalancing automates a process that for most investors is counterintuitive: selling some of your “hot” asset categories and moving the proceeds to the “cool” categories. Scientifically this typically gets you lower risk and higher returns, although most investors cannot emotionally bring themselves to make this type of switch. Think about it - what most people do is quite the opposite. They take money out of the cool categories and move it to the hot categories. This strategy usually provides them with the opposite of the desired result: higher risk and lower returns.
Since inception nearly two years ago, QAI has risen 7.6% while the S&P 500 has risen 62% in the same span. But as you note, it's not so much about garnering one-time outstanding returns but to limit risk; I suppose it's reasonable to think that if QAI had existed a little earlier it would have outperformed the broader market through the downturn? And that that's the point, balancing underperformance in bull markets with outperformance in down markets?
You answered your own question fairly well. I would add that although QAI did not exist prior to March of 2009, the IQ Hedge Multi-Strategy Index that QAI mimics did exist, and in 2008 it was down 11.12% vs. the S&P 500 being down 37%. This is a completely different asset class that provides low correlation benefits, so it is imperative to look at full market cycles as opposed to individual years or market cycle fragments. IndexIQ data suggests that during the period from January 1, 2005 through December 31, 2010 the IQ Hedge Multi-Strategy Index had a compounded annual return of 4.39% vs. the S&P 500’s compounded annual return of 2.29%, and that was achieved with a standard deviation of 7.41, which was less than half of the S&P 500’s 17.67.
Why might QAI not perform as well as you expect it to?
If the Multi-Strategy Hedge Fund AI asset class does not perform well on an absolute basis, that is not a concern of ours. Every asset class has good and bad years and as strategic investors Empirical Asset Management would not react to that data. If, however, IndexIQ’s replication process does not provide us risk and returns that reasonably match the Multi-Strategy Hedge Fund Index, whether that performance is better or worse than the index, then we would be concerned.
Our primary objective is to perform in line with the index, both in terms of risk and return. Although the portfolio has only been active since early 2009, we believe that the testing back to 2002 of their transparent and rules-based process provides reliable, fully vetted investment results for their replication strategy. The test of time and actual active management will provide the true answer to that question, but for now, so far so good.
Thanks, Mark, for sharing your process with us today.
Disclosure: Long QAI.
If you are a fund manager and interested in doing an interview with us on just one stock or ETF position you'd hold, please email us at Just One Stock.