Introducing Optimized Yield-Target Closed-End Fund Portfolios

by: ADS Analytics
Summary

A systematic investment approach has often been overlooked but it borrows the best from active and passive styles while avoiding their worst aspects.

We apply a systematic approach to constructing target-yield closed-end fund portfolios which are optimized for minimal volatility.

We believe minimizing volatility is important because it helps to mitigate behavioral biases and investment mistakes as well as preserves capital to be deployed at opportune times.

Our optimized portfolios look attractive relative to baskets of funds trading at a similar yield as well as popular closed-end funds of funds.

The passive vs. active debate that is raging on Wall Street often misses an important third leg of the stool - a systematic or "smart beta" approach. This sort of quantitative method of investing is still relatively small at $1trn of AUM, however it is growing quickly. Another reason it has not achieved as much prominence is that it shares characteristics of both passive and active styles - it is active in a sense that it deviates from any established static index but it is passive in a sense that it follows a set of pre-defined rules.

Our own approach to the income fund space borrows from this third style of investing. We do not follow a passive approach because of well-publicized issues with doing so for income assets. We also do not advocate a fully active approach because of a number of behavioral biases that such an approach opens up as well as an unfortunate lack of a crystal ball. Instead, we are adherents of a systematic and transparent investing style which we think takes the best of both worlds and avoids their worst characteristics.

In this article we take a systematic approach to a common problem for income investors - how to identify an efficient yield-target portfolio. By efficient we mean optimized for minimal volatility. Why minimal volatility and not say, maximal return?

First, forecasting returns is practically impossible on a short-term basis with any confidence while forecasting volatility is significantly easier since, among other reasons, volatility is much more stable than returns.

Secondly, we target minimal volatility because lower volatility and drawdowns supports many other aspects of the investment cycle - it helps mitigate behavioral biases such as selling in a midst of a deep drawdown and it also helps risk-based and leveraged investors, that is, investors who like to maximize yield per unit of risk and those who may want to leverage their portfolios to achieve a higher yield while controlling overall portfolio risk. Finally, it preserves capital to be deployed into attractive opportunities in case of a large risk-off period such as the one we saw this past December.

It's important to say that these portfolios, as all systematic strategies are constructed with a backward perspective, that is, there is an attempt in some way to forecast the future by looking on the past experience. Nothing in this process is guaranteed to work, however, we are of the view that if we focus on things that are easier to forecast (volatility rather than returns) and take a systematic and rigorous approach (rather than a purely discretionary, narrative-based star-manager-like approach) then we stand a higher chance of success. This is not the case just because of the "process over person" approach but also because a transparent process increases the chance of stick-to-itiveness and mitigates a lot of the behavioral biases and investor mistakes that usually go together with following purely discretionary active investment approaches such as selling at the bottom or chasing outperforming managers just as their performance is about to mean-revert.

Setting The Stage

What do we mean by optimized yield-target portfolios? As we suggested above, we optimize for two things: yield-target and volatility. More specifically, we define our two key parameters as follows:

  • Yield-Target: past 12-month distribution rate
  • Volatility: past 3-year daily gross price return volatility

There are two common definitions of distribution rates or, more commonly, yield. There is the past-12 month rate used here which is the sum of all distributions made in the past 12 months divided by current price and the indicative rate which is the most recent annualized distribution divided by current price. There are pros and cons in using either - using the past-12 month rate can overstate the rate for those funds that have cut their distributions in the past year however using the indicative rate can understate the rate for those funds that typically pay special distributions.

There have been two trends that have been responsible for driving a wedge between the two distribution rates in the recent past, first is the common tendency on the part of some funds to overdistribute and secondly, the rising cost of leverage due to Fed Funds rate increases over the last few years. Now that fewer funds are overdistributing and the Fed hiking journey is close to its end, we think using the past 12-month rate should be less controversial and probably more defensible than using the indicative rate.

We use a 3-year volatility rather than the last 1-year to obtain more stable and robust estimates that cover a longer trading period. There is a balance to achieve here - a longer volatility period covers more market environments at the expense of a lower fund universe. So we would not necessarily want to jack up the period to the last ten years as that would result in a substantially smaller number of funds and likely less robust portfolios.

The fund universe in question is our standard set of largely income-focused CEFs. We mostly exclude equity CEFs (mostly, because we include the covered call sector) and mostly include non-equity CEFs. We also exclude funds with less than three years of trading history so that our volatility estimates are more robust.

This is a snapshot of the latest yields and volatilities of our CEF universe.

Source: ADS Analytics LLC, Bloomberg

We also set a cap of 7.5% weight for any individual fund to ensure sufficient diversification in each portfolio.

How It Works

How does the portfolio optimization process work?

In order to minimize volatility for a given target yield, the optimization process relies on two metrics:

  1. Fund volatilities - specifically, the 3-year period we touch on above.
  2. Fund correlations - specifically, a fund's pairwise correlations to all other funds.

All else equal, the optimizer will select funds with lower volatilities.

However, generally speaking, all else is not equal and there will be many instances when a fund with a higher volatility will be negatively correlated to the other funds in the portfolio, meaning that on a portfolio basis, a higher-volatility fund will actually be a better choice than a positively-correlated lower-volatility fund.

There is wide variation in intra-sector correlations and this is what the optimizer takes advantage of. For example, municipals and mortgages generally have low correlations to other sectors (these are the lines with lighter-colored areas). There are also pockets of high sector correlations such as the one in the lower right corner in the chart below which has equity-linked sectors. This means the optimizer will avoid filling the portfolio with many equity-linked sectors and will likely add the right amount of low-correlated sectors such as municipals and mortgages. This does not mean however that only low correlated sectors are chosen by the optimizer as there are market environments where equities will outperform while municipals and mortgages will underperform.

Source: ADS Analytics LLC, Bloomberg

To concretely illustrate the power of diversification or as Harry Markowitz called it "the only free lunch" in investing, we take the optimized 8% target-yield portfolio and compare its volatility to the volatilities of funds that yield in the neighborhood of 8% (i.e. those yielding between 7.5% and 8.5%).

Statistic Volatility
8% Neighborhood Yield - Mean 12.9%
8% Neighborhood Yield - Median 12.6%
8% Neighborhood Yield - Minimum 7.4%
8% Optimized Target-Yield Portfolio 3.3%

Source: ADS Analytics LLC, Bloomberg

The table shows that even if we pick the lowest-volatility fund with a yield around our target (which has volatility of 7.4%), we still get a 42% reduction in volatility (from 7.4% to 3.3%) if we go with our 8% target-yield optimized portfolio in addition to the huge diversification benefit by going with a portfolio over a single-fund investment.

Introducing the Portfolios

There are a few different ways to get an intuition about the resulting optimized portfolios. In this section we compare our portfolios to other portfolio allocations and see how they fare.

Optimized Portfolios vs Individual Funds

In the chart below we plot all the funds in our universe (in blue) versus our optimized portfolios (in red). The yield goes on the y-axis and the volatility goes on the x-axis. The chart clearly shows that for a given yield (moving from left to right), the optimized portfolios volatilities are significantly lower than any individual fund with a similar yield.

Source: ADS Analytics LLC, Bloomberg

Optimized Portfolios vs "Neighborhood-Yield" Portfolios

As in our example above, we defined a "Neighborhood-Yield" portfolio as an equally-weighed portfolio of funds near a given yield-target. So, for example, an 8% "Neighborhood-Yield" portfolio is a portfolio of funds with yields between 7.5% and 8.5%. If we compare these portfolios to our optimized portfolios, we can see that there is a huge volatility reduction in the optimized portfolios.

Source: ADS Analytics LLC, Bloomberg

This is the case for two reasons as we described above: the optimizer tends to pick lower volatility funds and it tends to pick a portfolio where funds have low correlations to each other. The chart below illustrates the much lower average pairwise correlation of the optimized portfolio (in blue) versus that of the Neighborhood-Yield portfolio (in green) for an 8% target yield. It is especially interesting to see what happened in the risk-off period of December-2018 where the Neighborhood-Yield portfolio pairwise correlation shot up over 40% while that of the optimized portfolio increased less than 15%.

Source: ADS Analytics LLC, Bloomberg

Optimized Portfolios vs Funds of Funds

Here we compare our portfolios to popular funds of closed-end funds traded in the market. We have to admit upfront that this is not an entirely fair approach because the goals of our portfolios and these funds are different. It's not that we've come up with a better approach to funds of funds - it's that we are offering something different.

Our focus is on minimizing volatility while maintaining yield whereas these funds are after yield as well as capital gains. The management of these funds want to outpeform the market whereas that is not an explicit goal of our portfolios.

So our message here is not that you should sell any of these funds if you hold them. Our message is that if you buy these funds because you are targeting a certain level of yield, there may be a better way of doing this and you will save additional management fees in the process.

Source: ADS Analytics LLC, Bloomberg

Tire Kicking Our Porfolios

There are a couple of sense-checks we can do to check the robustness of our portfolios.

First, we should expect portfolio volatilities to remain monotonic, that is always rising as a function of target yield. In other words we should not find a 10% target-yield portfolio to have a lower realized volatility than a 7% target-yield portfolio, at least not for any reasonable length of time. We do this by taking our current portfolios and check their volatilities in the past. Although we only have two years of data (because of the 3-year limit on fund price histories which results in a 2-year history of 1Y volatility), the results look pretty good - the volatilities of the optimized portfolios do not cross over each other.

Source: ADS Analytics LLC, Bloomberg

Secondly, we can check on the average pairwise correlation of all funds in the portfolio. It is important for this metric to remain relatively stable because it ensures that portfolio volatility remains contained even if market volatility rises substantially. If the pairwise correlation is not stable, then we should have less confidence in the ability of portfolio funds to provide diversification. Checking the correlation on the 8% target-yield portfolio we can see that it generally remains stable, between 5% and 30% although there are bumps along the way which is not unreasonable given that asset correlations do tend to ebb and flow through different market environments.

Source: ADS Analytics LLC, Bloomberg

These two sense-checks gives us additional confidence that these portfolios should behave more or less as expected in the future.

Conclusion

Target-yield portfolios are a common offer of any number of investment gurus. These offers usually rely on three things: a macro-economic rationale, a sense of relative attractiveness of various funds and their future relative performance plus a view of the overall direction of the market. Our approach is different - our view is that it is difficult to get all three of the above right consistently, if ever. So instead we take a quantitative tack - not only because we think it is more reliable and transparent but also because minimizing volatility is a virtue in itself. It helps to mitigate behavioral errors, preserves capital to be deployed at critical times and can support greater portfolio leverage. In a nutshell, we think in an environment of greater uncertainty and volatility, these portfolios can be a defensive tool for different investor types and investment styles without compromising on income.

In the coming weeks we plan to launch Systematic Income - our Marketplace service on this platform. In addition to detailed analytics of CEF funds and sectors, frequent tactical screens and ideas, we plan to publish and discuss regular updates and performance of our target-yield portfolios. We hope you can join us.

We hope you enjoyed this article. If you would like to automatically receive similar analysis from us, please follow us above by clicking on the Follow label next to our name.

Disclaimer: This article is for information purposes only and does not constitute investment advice or an offer or the solicitation of an offer to buy or sell any securities. Past performance is not a guarantee and may not be repeated. Investment strategies are not suitable for everyone and you should always conduct your own research or speak to a financial advisor. Although information in this document has been obtained from sources believed to be reliable, ADS ANALYTICS LLC does not guarantee its accuracy or completeness and accept no liability for any direct or consequential losses arising from its use. ADS ANALYTICS LLC does not provide tax or legal advice. Any such taxpayer should seek advice based on the taxpayer’s particular circumstances from an independent tax advisor.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.