Your Guide to Private Money Managers - Money Manager Review
Money Manager Review 
For a Free Tour of our Manager Performance Rankings
Call 1-415-386-7111
Follow Us Facebook Twitter

    Guest Interview:

   Good Harbor Financial, LLC

    155 N Wacker Drive, Suite 850
    Chicago,IL 60606

    Telephone: 312-224-8150
    Fax: 312-674-7230


    Interview Quarter: 3Q2011

 Neil Peplinski, CFA,

 Co-Founder & Managing Partner,

  Neil Peplinski, CFA, is co-founder and managing partner of Good Harbor Financial, an SEC registered investment advisor firm focused on tactical asset allocation. Previously he served as portfolio manager and quantitative analyst for Allstate Investments overseeing a $400 million dollar book of structured assets. Neil earned an MBA with High Honors from the University of Chicago Graduate School of Business. He also holds an MSEE in electromagnetics from The University of Michigan and a BSEE in electromagnetics from Michigan Technological University where he graduated summa cum laude.  
  Why did you adopt tactical allocation model in 2003?  
  2003 is the first year we started managing money under Good Harbor Financial using our tactical asset allocation strategy. However, I had actually been using active approaches for a number of years. In fact the Good Harbor model is based on the research I did back in the late 90’s while searching for an alternative to buy and hold. While I’ve come to appreciate this more over the years, even back then the idea markets are efficient and therefore active management cannot offer value just never sat well with me. It’s not so much that I believe markets are inefficient; in fact as a graduate of the University of Chicago’s Graduate School of Business I take the efficient market hypothesis very seriously. However, I don’t think the ability to make market forecasts and the existence of market efficiency necessarily need to be in conflict. Take the random walk argument for example. Whether it is appreciated or not, those who argue markets are efficient because prices follow a random walk are making a subtle but important assumption – namely that expected returns do not vary with time. That is akin to saying investors felt the same about equity risk in April of 1999 as they did in October in 2008. I am willing to bet that was not the case. In fact, studies show not only do expected returns change over time, they in fact account for a majority of stock market variations. As such, if people’s views towards risk (i.e. expected returns) are slow moving and investors dislike risk in bad times (i.e. they demand higher returns during recessions), then there may well be times when it is more favorable to hold stocks and times when it is more favorable to have less equity exposure. These changing risk premiums do not represent an arbitrage opportunity or a means to quick guaranteed profits. But the understanding that they vary can be useful in positioning a portfolio tactically during certain market environments as a means to reduce losses or capitalize on potential gains. This can all be garnered from a straight-forward discounted dividend pricing model, so we’re not dealing with esoteric or complicated theories here.  
  What do you think about buy and hold approach?  
  One of the main arguments in support of buy and hold is the notion investors cannot afford to miss the best market days. However, what is lost in that logic is the fact buy and hold investors will also never miss any losses. Despite our best wishes, the markets don’t just go up, so losses and bear markets are something every investor must deal with. Consider the NASDAQ. On March 10th, 2000 the NASDAQ composite reached a high of 5048.62. Ten-years later, the index is still down over 50%. Unfortunately, this isn’t the only “lost-decade”. From December 1968 to August 1982, the S&P 500 Index measured a loss of 1.18%. In other words, after a fourteen-year commitment to the stock market, you had less money than when you started. Even worse, from March of 1999 through March of 2009, a period now all too familiar to present-day investors, the S&P 500 posted a loss of nearly 40%. These are buy-and-hold returns. And they can be hard to stomach. If a typical investment horizon is twenty years, a ten-year drought consumes half of the available time for making money. So if ten years in, we’re break-even (not to mention possibly down 40%), an 8% annual growth assumption for equities now needs to be bumped up to 16% per year just to stay on track, not including the effects of compounding, which only work to make matters worse. And while it can be easy to dismiss a year like 2008 as an outlier, the fact is market declines of 20% or more happen more often than we might care to admit. Since 1900 the S&P 500 index has seen declines of 20% or more fourteen times. So on average, we can expect some significant market turbulence about every eight years. And once the decline is set in motion, it takes on average 6.5 years to reach a new equity peak (think of what that means for the recovery from 2008. It could very well be 2015 before we see a new equity peak). With the benefit of hindsight and a two hundred year data set it might be reasonable to conclude markets always come back. But the concern isn’t whether prices come back. It’s do they come back within an investor’s time frame? Losses to an investment portfolio are damaging in two regards. The first is simply the loss in value. The second is the return required to recover. Recall a 50% loss requires a 100% return to get back to break-even. To my knowledge, there is no reason to believe a 100% return is more likely to occur immediately following a 50% loss. In other words, the act of losing money does not make getting it back any easier! Therefore it seems worthwhile to not only attempt to capture market gains but also have some mechanism to at least attempt to minimize losses during sustained bear markets. By very definition this requires some sort of active management strategy.  
  How do you select your portfolio investments?  
  At Good Harbor Financial we use a fully objective asset allocation model that analyzes three main categories of information; momentum measures; yield curve dynamics and economic conditions. The momentum measures are designed to provide a read on strength in stock and bond prices, and the relative strength between them. Similarly monitoring the U.S. treasury yield curve can provide insights into investor sentiment and aggregate demand for capital. As the economy starts to enter a difficult period, changes in the shape and levels of treasury yields, whether through government policy intervention or classic flight to quality effects, can be valuable information. Economic data designed to measure the output level and growth rate of the U.S. economy further complete the picture by providing direct business cycle readings. The main thesis involves assessing the direction of the risk premiums mentioned above. If investors are becoming less risk tolerant, we ultimately expect further pressure on equity prices and therefore want to position the portfolio more defensively. Likewise if investors are becoming more risk tolerant, we’re going to allocate more heavily to equities.

Regarding investment vehicles, the universe of tickers we use is actually quite small. With some reflection, it should be clear the idea of changing risk premiums is a broad-based concept. Investor return expectations certainly apply to individual stocks, though if we go this route we have idiosyncratic (i.e. company specific) risk to deal with as well. For our tactical asset allocation strategy, what we’re most interested in is whether investors as a whole are feeling better or worse about the stock market as a whole. Therefore it should make sense that when we want equity exposure in a portfolio, we choose a broad-based stock market index product, such as an S&P 500 or Russell 2000 exchange trade fund (ETF) or mutual fund. When we want to be defensive with a portfolio, however, there are three basic choices – cash, inverse stock market funds or U.S. treasuries. The benefits of cash are obvious – no losses. The risk of inverse stock market funds should be equally clear. If we’re effectively short the equity market (by being long an inverse ETF, for example) and the market rallies, it can be costly. Certainly if we have a tactical strategy that is highly accurate, shorting the market when we want to be defensive would invariably be the right place to be. But in the business of active investment management we’re happy to be right 60% of the time. Therefore we leave inverse ETFs for the gunslingers. This leaves U.S. treasuries. The benefits include potential for yield above cash. However, unlike cash, there is also the potential for loss. Our research suggest, though, the benefits outweigh the drawbacks. By examining past economic downturns, we find a 10-year bond fund has returned on average 1.55% a month during recessions. Furthermore we find bonds outperform stocks during both the front half and back half of recessions, on average. As a tactical manager this is good news as it means we don’t need to be exact in our assessment of where we are in the business cycle. Therefore when we’re looking to be defensive with the portfolio our allocation will be to products tracking U.S. treasury rates.

It is important to note we expect this to be the case even in low interest rate regimes. When we are on defense, we’re not looking to secure yield by holding bonds until maturity. We are not making a long-term bet on rates. Instead, we are looking to capture the capital gain that comes from changing yield levels. As long as rates are greater than zero, we expect U.S. treasuries to provide some semblance of a safe-haven when equity markets drop. Just look at the fourth quarter of 2008. At the end of November the 10-year treasury rate was just below 3%. It was hard to imagine rates going any lower, but just three weeks later rates were closing in on the 2% level. If nominal interest rates ever landed on zero, then clearly cash becomes the only logical choice – why risk the downside with no potential upside. But I don’t expect that to happen.
  How did you avoid large losses in 2008 market meltdown?  
  One of the main benefits of using a model based approach is the ability to remove emotion from the actual investment decision making process. This can be key in environments like 2008 or even 2009. As 2008 unfolded, we saw weakness in several of the data inputs we use. Stock price momentum was weakening. The yield curve is flattening. The U.S. economic output level is dropping and we’re seeing a negative growth rate. In general it is pretty easy to sit back and assess the situation as weak at best. Our tactical model, however, allowed us to take this one step further as it flagged a fully defensive portfolio allocation at the end of September and continued to have zero weight to equities for the remainder of the year. That obviously helped minimize losses. But as I mentioned earlier, when we’re looking to be defensive, we take our allocation in U.S. treasury funds. The rally in treasuries in the fourth quarter of 2008 provided a nice boost to the portfolio as well. This highlights the point earlier that bonds can be a nice alternative to cash during market meltdowns.

The important observation here is that we had a strategy that was specific in navigating the turbulent times in the markets. We know plenty of money managers who definitely felt the weakness during 2008. But without a specific mechanism to help make actual decisions, most took little or no action. If the economic backdrop, sensational headlines, and constant bankruptcy news paralyzes decision making, then you’re at the mercy of the markets. And it gets harder as the sell-off continues. To tell clients to sell after a 10% decline is hard. It’s even harder after a 20% or 30% drop.

In a similar fashion, we started to see positive signs early in 2009 with additional confirmation in March of that year. As such, we were partially allocated to equities in the first quarter and then fully allocated to equities starting in April. We held that position for most of the year. So we not only avoided losses in 2008, but followed up with a solid performance in 2009. Having a mechanism to take equity risk when the odds are favorable is the other side of the same coin. And while 2008 and 2009 were perhaps unique in the magnitude of the moves, similar type of market dynamics has occurred repeatedly during the past. This is effectively what we try to capture with our tactical allocation approach.
  When does tactical allocation perform best?  
  This is obviously very strategy dependent. In general, though, tactical approaches will fall into one of two categories; momentum-based and mean-reverting. Approaches having a momentum component will naturally perform best when the markets are trending in either direction. Conversely, mean-reverting strategies tend to perform best in sideways, high volatility environments.

The Good Harbor tactical model assumes changes in investor risk premiums are persistent. Therefore we tend to perform well when markets trend, in either direction. With any active strategy a compromise must be made between being so quick to react to changing information that market noise becomes a problem, and being so slow to rebalance that we miss the very moves we’re trying to capture. In developing the model I spent a great deal of time on this issue. In the end, the research suggests a four-week hold period strikes a good compromise between the two extremes. As a result, we rebalance the portfolio approximately monthly. This means we will tend to get whipsawed in years like 2005, where the markets move more or less sideways month to month and there’s no real conviction the data series we monitor.
  How do you manage portfolio risk in a volatile market?  
  We don’t manage risk during volatile markets any differently than in non-volatile markets, at least in concept. One of the other advantages of a model based approach is the benefit of analyzing a strategy’s performance during different market environments. By doing this we can formulate our risk management plan ahead of time and then execute it right along-side security selection. In the case of our tactical allocation model, we are comfortable with the expected risk-adjusted performance and therefore make no adjustments if market volatility rises. Given specific portfolio allocation constraints we expect this strategy to have volatility levels that are lower than the overall equity markets on a relative basis across differing market environments.

For our portfolios that utilize leverage, we attempt to match overall portfolio volatility to the S&P 500 index volatility using a three-year trailing window. This suggests if the portfolio volatility rises relative to the index, we will reduce the leverage target. This is also something we were able to thoroughly analyze given the context of a fully objective model-based framework.
  How did you manage the psychology of your team in the crisis time?  
  I alluded to this earlier. One of the main benefits of a model based approach is the ability to manage the decision making process during difficult market environments. Therefore if you maintain the discipline to follow the strategy during crises, the psychology takes care of itself. However, this isn’t always the case. And it’s easy to underestimate the effort it takes to maintain that discipline. For a model based approach like we use at Good Harbor Financial, the times the psychology can be difficult is naturally when performance is lagging the benchmark. It’s during these times that we rely heavily on our experiences as well as the back-test results. No active strategy is right all the time. So when things seem like they are going off track, it’s important to assess whether you believe your strategy is no longer valid or whether it is simply the inevitable drawdown period. By having a wealth of historical data, we can better judge if we are in a normal loss environment or whether there are real concerns about the approach itself. For all of our strategies, not just the Good Harbor tactical allocation model, we continually review the out-of-sample results to the in-sample numbers. This allows us to assess, from a statistical perspective, whether the strategy is behaving as expected. If it isn’t, we review matters further to understand why. In extreme cases this may ultimately mean retiring an approach.

We all know past performance does not guarantee future results (although ironically it is often all we have to work with). It could be the future will look so radically different from the past that any given approach ceases to provide value. Even buy-and-hold can’t escape this possibility. However, if an active investment strategy is well conceived, is based on underlying economic principals and has a decent track record, hypothetical or real-time, it is worth consideration in my opinion. Effective active managers will exercise judgment and expertise to improve a strategy’s odds of performing in the future. This goes for those managers who favor model-based approaches as well as those who focus on fundamental research. Beyond that, we’re at the mercy of price action and the whims of the markets!
  What are some of your hedging techniques and risk management rules?  
  For the Good Harbor tactical allocation model, there is no explicit hedging. Recall the point of the model is to try and minimize losses during sustained market down turns. We do that by allocating the portfolio to U.S. treasury based exchange traded funds (ETF) and mutual funds, when the model calls for it. Because the investment vehicles we use are so highly liquid, if we want equity exposure, we put it on. If we want risk off the table, we sell equities and buy bond-based products. The capacity of such an index based strategy is very high. So other than ensuring efficient trade execution through relationships with key ETF liquidity providers, there is no need to consider other hedging tools. This is the direct result of the tactical model being based on broad U.S. market sectors. The situation is different for some of our other strategies, where we implement options and futures as the main tools to hedge various components of risk.

For the tactical portfolios utilizing leveraged funds, we limit the maximum leverage allowed. We do this even if the volatility matching approach outlined previously would allow for higher levels. This helps reduce (although not completely eliminate) the impacts of being caught in an equity volatility spike.
  When do you consider active management is not an efficient strategy?  
  The simple answer is when it is not working. However, determining when a strategy is not working is as much a philosophical discussion as anything else. Every investment strategy whether active or passive, is going to experience losses, drawdown, periods of underperformance, etc. However, one advantage of a passive strategy is if things are going south we know it’s not the manager’s fault. While I believe active management can add value, I’m not saying all active management is a value added proposition. This is the biggest reason we tend toward model based approaches. Doing so allows us to set our expectations going in and then monitor whether those expectations are being met. If they are, we forge ahead. If not, we reevaluate.  
  What did you learn from recent crisis (2008)?  
  This may sound odd, but at least from the perspective of our tactical allocation strategy, 2008 was not that unusual. It was certainly dramatic in the magnitude of the drop in equities. But the dynamic at work is something we see repeatedly in our back-test data. The same can be said for the subsequent rally in 2009.

Probably the greatest impact 2008 had for us was the increased acceptance of tactical asset management that resulted. We’ve been using a tactical framework from the beginning. However there is no denying there was less interest prior to the crisis. Many investors forgot about the 2000 to 2002 sell-off and were feeling good about the equity rally from 2002 to 2007. But this latest episode reopened a lot of eyes to the fact that market declines are something that deserve attention. It’s like I mentioned earlier, essentially having a negative return over a ten year stretch, although evident in historical data, was something most present-day investors hadn’t anticipated yet alone prepared for. We saw signs of that psychology even in our own clients. From day one we had been describing the objective of the Good Harbor tactical model as trying to capture equity gains when markets are rising, but have a defensive mechanism should we enter a multi-month market decline. But for the first five years we managed external money, the markets essentially only went up. Some investors questioned the value of a strategy that was effectively producing market returns. Fortunately our clients stayed patient. Now with 2008 in the books, they get it. Needless to say our original client base doesn’t call us anymore. But the fact it took a 2008 type of crash to drive the point home certainly reflects the human nature element to investing. This is something that will never go away. And quite frankly I wouldn’t want it to change. It’s this human element that ultimately drives the changes in risk premiums upon which our tactical allocation is based!
Manager Services | Site Tour | Site Map | About Us | Interviews | Contact Us
© 2014 Money Manager Review TM. All Rights Reserved. Terms of Service | Disclosure & Privacy Statement1.