This post is a continuation of a series where we will be providing some of our own thoughts and commentary the conversations we had in the first season of our new podcast.
This post covers our conversation with Eric Ervin, which you can listen to here.
3:12 - Eric talks about his background as a financial advisor and trying to build absolute return strategies for his clients.
Justin Sibears ("JS"): I think Eric’s experience with liquid alternatives mirrors that of many other advisors. Many mutual fund or ETF-based alternatives offerings are simply a watered-down version of true alternatives due to constraints around leverage and shorting. Alternatives that have too little volatility can struggle to be anything more than, as Eric puts it, “high priced T-bills.”
Nathan Faber ("NF”): Having the background of being a financial advisor helped Eric understand what was missing in the marketplace once he transitioned to asset management. His experience is unique in that many clients do not have the capacity to stomach a 50% alts sleeve. However, for those who can hold that many alts, there is a lot of opportunity to pursue unique goals that aren’t as dependent on the market (see our previous research on building an unconstrained sleeve).
Corey Hoffstein ("CH"): I was particularly excited to talk about Eric not only because we had written about DIVY in the past, but because I knew some of his background using a significant amount of alternatives in his role as a financial advisor. Eric does not just preach using alternatives: he lived it with his clients. So he has a deep appreciation of how to think about evaluating alternatives, incorporating them into a client plan, and dealing with behavioral issues they can cause.
5:53 - Eric introduces his mental framework of breaking down the alternatives universe into "risk reducers" and "return enhancers."
JS: I love this framework, especially because it can be very helpful in terms of driving home client expectations.
NF: We have used this framework before to bucket tactical strategies. I love his line about a strategy becoming a risk enhancer/return reducer.
CH: I think the simplicity of this framework is what makes it powerful. There is plenty of room to add complexity further down the decision tree, but starting with this split allows us to re-frame why we hold something in our portfolio.
7:22 - The importance of understanding the difference between assets and strategies in the alternatives space.
JS: I tend to believe that alternative strategies tend to be best kept in private vehicles like hedge funds due to the flexibility it affords. Alternatives that simply offer exposure to a unique asset class can often be structured in a much more traditional way, barring hurdles like liquidity.
CH: I think this is a really important distinction and it is important for investors to think about what they are actually accessing. For example, "smart beta" products are designed to give investors access to both an asset class (typically equities) and a strategy (e.g. value, momentum, dividend growth, etc). (See It's Long/Short Portfolios All the Way Down). By and large, however, it is the asset class that drives performance. In alternatives, however, a more pure long/short implementation may isolate the strategy exposure. So while the investor may technically hold equities, the return profile may not look equity-like at all.
Trying to figure out what a manager is really giving you is important for performing due diligence and determining what you are actually paying for.
9:39 - Addressing the behavioral issues of placing 50-60% of a client portfolio into alternatives.
JS: Long-term success of alternatives really requires thinking about them like investors think about bonds. Intuitively, most investors understand that they hold bonds for the diversification benefit and that they will behave very differently from stocks. As a result, you don’t see investors rushing out of bonds when equities are up 20% and bonds only return 3%. Alternatives need to be thought of in a very similar way. If you are going to ditch alts as soon as they underperform equities, you probably shouldn’t even bother in the first place.
NF: This amount of alts can provide meaningful diversification, but it is key to remember that diversification will always be disappointing. An ideal alts bucket would have a very asymmetric payoff profile that provides good diversification with equities in down years. However, when equities are up a lot, alts likely will not be up as much. As long as appropriate expectations are set from the start, this amount of alts can provide a unique return profile over the long-run for an investor.
CH: I think an important comment here is when Eric talks about having an allocation that is meaningful enough to actually offer meaningful diversification.
If we're trying to be different than the market, we have to invest differently.
It's a painfully simple concept, but getting investors to buy into it is not easy.
12:13 - Shorting the subprime market and building conviction in the trade.
NF: It’s always fun to hear stories about winning trades, and I like the depth with which Eric tried to understand what he was investing in. If more investors took time to do this (which will never happen!), the subprime crisis might not have been so bad in the first place.
CH: I think there are two really important threads of thought worth considering here. The first is Eric's willingness to get his hands dirty and really understand the trade. It was not enough for him to simply follow the guidance of a portfolio team he trusted; he went so far as to learn to price credit default swaps himself. I think that is an important lesson for anyone looking to get involved in the alternatives space.
The second line of thought that I think is worth addressing is that for every Eric Ervin, there are probably a dozen other investors who become equally educated and invested in a narrative that did not play out (e.g. listen to our conversation with Adam Butler). In hindsight, Eric's trade was "obvious"; it was "just the math." But it's worth considering whether Eric's takeaway lesson from the situation would be different had the credit crisis not unfolded the way it did.
15:20 - Eric's approach to building a portfolio out of alternative investments.
JS: Using a building block approach as Eric described can really help to simplify the thought process around using alternatives.
NF: Classifying investments based on how they are likely to behave is a good way to keep a portfolio in line with expectations. Eric’s example of leveraged fixed income being a return enhancer illustrates how we need to move beyond simple asset class definitions when dealing with alternatives.
CH: What intrigued me here was the juxtaposition to the depth with which Eric was willing to dive to understand a strategy, but the simplicity of the mental framework he used to put the pieces together. In a quantitative sense, this simplicity may be a heuristic that helps minimize estimation error (likely a "high bias, low variance" trade-off) that might otherwise be incurred if he tried to use a more complex optimization-driven approach.
19:42 - The red flags Eric has identified in the alternative space.
JS: Overdiversification is an interesting topic and can be a bit misleading as a term. We want as many uncorrelated bets as possible. If you give me a million uncorrelated bets with positive expected returns, you better believe I’m allocating capital to all of them. There is no virtue in trying to pick 10 or 20 of the million. All you are doing by going down this path is diluting your risk-adjusted return. Where overdiversification does become a problem is when it starts to dilute the expected return side of the equation. In addition, diversification does not relieve you of due diligence responsibilities. Warren Buffett famously said, “Diversification is protection against ignorance.” Personally, I would change this slightly to say, “Don’t use diversification as protection against ignorance.” Diversification is great if I’m allocating to a set of strategies or asset classes that I have high confidence in.
Now, this all assumes that I can properly employ leverage to get the desired volatility level. If I can’t, I may be better off using a less diversified portfolio even though the result will be far from optimal.
NF: I understand Eric’s point on diversification as “dilution”… going back to his high-fee T-bill take on alternatives. Having high volatility, pure alternative strategies gives investors the building blocks to build their own alts sleeve based on their risk tolerance and investment objectives. The problem comes in when we are overpaying for what turns out to be a low risk bond-like return profile that doesn’t have enough juice to really diversify equity positions.
22:07 - The launch of Reality Shares in 2011 and the Dividend Swap ETF, DIVY.
NF: I learned about DIVY in 2015, shortly after its launch. What a unique ETF! It was one of the first where I had to spend a decent amount of time figuring out what it was doing. I wish I had this podcast at that time.
CH: The ETF and mutual fund graveyards are littered with many well-intentioned launches. I think what is unique here is that Eric had the perspective of an advisor who was actively using alternatives, allowing him to think of product ideas that were both (1) missing from the market and (2) were ideas he would be willing to use.
24:26 - A basic overview of swaps in general, and more specifically dividend swaps.
NF: Like most derivatives, swaps are a way to transfer targeted risks. Dividend swaps allow you to separate dividend risk from the total return.
27:02 - Corey takes a step back from the discussion to discuss his mental model for exploring new ideas by thinking about the extremes of the problem.
JS: A perpetual dividend swap is equivalent to holding the equity itself. Therefore, if we held a combination of 0-5 year dividends swaps and 5+ year dividend swaps, we should earn in aggregate the equity risk premium. By doing this decomposition, we can then start to think about where we earn more of the risk premium. My intuition, which could be wrong, is that the short-end should deliver more premium relative to the risk taken on because these cash flows may be more sensitive to a market crash. The global financial crisis had a major impact on short-term dividends, but likely will end up having little impact on dividends paid twenty or thirty years down the line.
CH: The idea behind this framework is, basically, that things are often easier to understand in specific examples rather than on average. By focusing on the extremes, we can identify the constrained, limiting behavior and then just use simple linear interpolation to try to get an understanding of the middle. It's certainly not a perfect framework, but I like it as a starting point when I'm learning something new.
29:06 - What is the dividend risk premium?
NF: The dividend risk premium is higher when there is more uncertainty about the future dividends. It’s interesting to think about the difference in uncertainty between the sticky-dividend US market vs. the European market that is more comfortable with cutting dividends.
JS: The beta of the dividend swaps provides some nice intuition around where we are truly bearing risk when purchasing equities. If the beta of a series of 0-10 year dividend swaps is 0.7, that basically means that when we buy stocks 70% of the risk is in the first ten years and 30% is after ten years. We would expect that the beta of the front piece will increase when the equity risk premium increases, since discounting at a higher rate will mean that our near-term cash flows make up a larger percentage of our overall market valuation.
CH: I'll admit that it took me a long time to really grok why there was a premium associated with dividend swaps. Not because it was particularly complicated, but because there are a large number of arguments. However, I think at the end of the day, if we apply Occam's razor, the risk argument is the cleanest and other associated arguments might just add some gravy from time-to-time (e.g. banks needing to offload realized dividend risk due to structural product issuance).
33:58 - Eric discusses expectations for the level of return and volatility associated with the dividend risk premium.
JS: One thing that is fascinating to me about dividend swaps is the role that company behavior plays in returns. The behavioral tendency of firms not wanting to cut dividends has a real impact. While this matters for equities generally, I would argue that the impact is less important (e.g. straining to pay dividends today may hinder the ability to pay tomorrow and the effects may then cancel out). Yet in the dividend swap market, the decision to pay dividends that the company cannot really afford may not impact investors at all (at least not until the investor rolls into new swaps).
NF: Higher interest rates should be beneficial to dividend swaps since you can then earn a higher rate on the cash collateral portion of the swap. Higher interest rates may also encourage companies to increase their dividends. If these increases occur at a faster rate than market expectations, then dividend swaps will also benefit. The volatility can come from the swap market pricing in higher dividend payouts that are not matched by the companies.
CH: To Justin's point, the culture of dividends in America has lead to a significantly lower dividend risk premium than what has been documented in Europe, where cutting a dividend is less taboo. While a culture of dividends may help enforce better corporate behavior, it's trivial to think of examples whereby cutting the dividend is actually the better long-term decision. A number of ETFs have come to market in recent years based upon this idea, recognizing that "years a dividend has consistently grown" may not be the best proxy for dividend health or stability going forward.
38:00 - If dividend expectations are well established a year out, why does the dividend risk premium exist?
JS: The economic rationale for premia in investing generally can be categorized into three buckets: behavioral, risk, and structural. Dividend swaps offer a premium that is part risk-based and part structural. The good news to hear is that even if the structural piece goes away (e.g. folks get comfortable enough with dividend swaps to the extent that demand equals supply), the risk piece will be there to provide some compensation.
NF: Aside from dividend swaps, this part highlights how retail investors using derivatives (e.g. structured notes and indexed annuities) often miss out on the dividends. As long as these products are still popular, the structural aspect of dividend swap demand will persist.
42:44 - How does the increasing scope of buybacks affect dividend swaps going forward?
NF: The beautiful thing about dividend swaps is that they can earn a premium despite lower dividend growth expectations in light of buybacks.
CH: I think this question is inextricably linked to the next, which is what makes dividend swaps interesting. If buybacks become preferential to dividends in the future, however, I do wonder what will happen to liquidity in the dividend market.
45:09 - Why growth in dividends can be negative but the dividend risk premium can still be positive.
JS: Markets move when new information is different from expectations. Bad economic data, for example, will not necessarily mean that the market will sell-off if it still happens to beat expectations.
NF: The big risk is always the big surprise, like dividend cuts during the Financial Crisis. However, I would expect that longer term dividend swaps would be more immune to these crisis events as expectations are generally much more pessimistic initially because of behavioral biases.
CH: I think this is a subtle nuance that is important to consider. The risk premium here is based around expectations. e.g. If the market expects a growth rate of -5% in the future, I would still demand a discount to that growth rate to insure it. So dividend swaps are really a pure play on expectations of growth and not the growth rate itself.
48:05 - Corey asks Eric to walk through his own framework for evaluating and incorporating alternatives with respect to DIVY.
JS: Dividend swaps will have some optionality in them from a risk perspective in the sense that market sensitivity will change depending on the level of the market, just like the delta of an option will change with the price of the underlying. We would expect dividend policy to be relatively unaffected by modest market declines, hence a lower beta in this environment. This will hold until economic conditions deteriorate so severely that actual ability to pay dividends is impacted (e.g. banks during the financial crisis) and betas rise.
NF: DIVY has had a low correlation to SPY and AGG. And Eric brings up a good point about positive correlation when you want it and negative correlation when you don’t. DIVY has had this relative to SPY with positive correlation in up months and negative correlation in down months.
CH: Like most risk premia, we would probably expect the return profile to have a non-linear relationship to equities. To Eric's point, the board of directors of a company probably will no consider a dividend cut simply because of a market decline. However, if the board is considering a cut, it is likely an environment where the company has suffered significantly and, therefore, the stock is suffering significantly. So we would expect a high degree of correlation in negative tail events. However, because these are swaps on the aggregate dividends of companies in the S&P 500, the swaps should be largely immune to idiosyncratic risks, but still sensitive to broad, economic risks.
56:12 - If you were an investment strategy, what investment strategy would you be and why?
NF: When many alternative strategies need to be higher volatility to have an impact on a portfolio, at least for the typical allocations that most investors hold, Eric’s personification as a lower volatility strategy diversifies his practice, as we saw for both Toby Carlisleand Adam Butler.