Talking Tacticals: Trading Opportunities In Closed-End Funds

by: ADS Analytics

The discount dynamic of closed-end funds present potential tactical trading opportunities for fund investors.

We explore the discount sector spread - an intra-sector valuation metric which historically has delivered high single-digit alpha.

In the preferred sector we overweight JPI and PFD and underweight HPS and PSF among others.

In our last Weekly, we discussed why we thought the current trading environment of heightened volatility was conducive to opportunistic tactical trading of closed-end funds. In this article we introduce a metric - discount sector spread percentile - that we think is particularly helpful in identifying trading opportunities. We plan to make it a regular feature of our reports as well as share quick-takes on the blog which are sent to our real-time followers.

According to the latest figures, our recommendations are as follows:


  • Nuveen Preferred and Income Term Fund: (JPI)
  • Flaherty & Crumrine Preferred Income Fund: (PFD)

  • Flaherty & Crumrine Total Return Fund: (FLC)
  • First Trust Intermediate Duration Preferred & Income Fund: (FPF)
  • Flaherty & Crumrine Dynamic Preferred & Income Fund: (DFP)


  • John Hancock Preferred Income Fund III: (HPS)
  • Cohen & Steers Select Preferred & Income Fund: (PSF)
  • John Hancock Preferred Income Fund: (HPI)
  • John Hancock Premium Dividend Fund: (PDT)
  • John Hancock Preferred Income Fund II: (HPF)

The table below captures the relevant data (as of 10-Jan) for the above recommendations. For each fund, it shows the current discount, the discount spread to sector average, the spread average and the percentile of the current spread. We have removed those funds with percentiles between 10% and 90% as they present less compelling opportunities. Funds we are overweight are in a green box and funds we are underweight are in a red box.

Source: ADS Analytics LLC, Bloomberg

Before going further, let's quickly describe the discount sector spread percentile by breaking it down into its components:

  • Discount - we are looking at the fund's discount, that is the percentage difference between its price and NAV
  • Sector - we are comparing the fund's discount to the sector discount
  • Spread - more specifically, we are taking the difference between the two where we define the spread as the fund discount less the sector discount. In other words, if the sector discount is -5% and the fund discount is -4% then the spread is -1%.
  • Percentile - we quantify the current spread relative to its history by calculating its percentile - 100% means the current spread is the highest in the rolling window and 0% means it is the lowest. The percentile calculation is very powerful because it not only gives us a way to talk about where the current spread is relative to a fund's history but also gives us the ability to compare one fund's valuation against another.

The upshot of this metric is that we think it makes sense to overweight funds that have low percentile readings and vice-versa. Low percentile readings of the discount sector spread suggests that the fund is trading at a wider-than-historic discount relative to the sector. The working assumption, which we will address below, is that funds that trade at a wider than historic discount relative to their sector tend to mean revert, thereby outperforming the sector in the near term.

It always helps to have a behavioral or intuitive explanation for any quantitative market strategy. For example, momentum has been shown to work across many periods and asset classes because investors tend to herd and use strong performance as a reason to buy and poor performance as a reason to sell which exacerbates existing price trends. Is there a behavioral explanation for mean reversion in discount sector spreads? We think so.

Closed-end funds have varying liquidity and their prices get pushed around by transaction flows to varying degrees. However, because closed-end fund investors tend to allocate to funds on the basis of distribution rates, among other factors, a relatively lower price of a given fund makes it more attractive since a lower price raises the fund's distribution rate and draws demand until the distribution rate falls back to its previous equilibrium. This, of course, does not always hold but it's a decent rule of thumb.

Hierarchy of Tactical Metrics

We are not strangers to tactical trading screens - we have published them regularly on the blog. We have focused on the standard metrics like prices, discounts, yields, volatility and others, sorting by highest or lowest to identify potential opportunities.

In discussing useful tactical metrics, it's worth to break them down into different types.

  • Absolute - e.g. widest discounts in CEF universe. This metric is certainly interesting but we think it does not give a lot of actionable information as certain types of funds e.g. low-yield muni or equity funds tend to persistently trade at the wider discount range of the CEF universe.
  • Relative-to-Self - e.g. a fund is trading at the widest discount in its history. This metric is an improvement on the absolute metric above however there are many scenarios where a fund can be trading at the wider range of its history such as during a large market sell-off e.g. during the financial crisis so just because a fund's discount is at its widest does not necessarily make it an interesting tactical opportunity.
  • Relative-to-Sector - e.g. a fund is trading at the widest discount spread relative to the sector discount in history. This metric solves the two problems above by focusing on the fund's relative rather than absolute valuation but also taking into account the performance of the fund's peers.

Hopefully the breakdown above makes it clear that the further down you go on the list, the more information is available to the investor and the better decision s/he can make. Over the last year we have framed our comments on the closed-end fund market by focusing more on fund sectors rather than individual funds. We think this allows a more coherent understanding of the market dynamics as well as provides a more robust portfolio construction process. And as this discussion suggests, it aids in recognizing tactical trading opportunities as well.

Discount Sector Spread Percentile - An Example

To provide more intuition around the discount sector spread percentile we go through an example in this section. We use the preferreds sector in the example for no major reason except that we've noticed a lot of interesting opportunities in the sector as it has been quite volatile due to the recession fears around the financial sector.

In the series of charts below we show how we calculate the discount sector spread percentile for a given fund and provide an intuition around its dynamics. Within the preferreds sector we choose to focus on the Nuveen Preferred and Income Term Fund (JPI) because it looks the most attractive on our screen.

First, we calculate the fund discount (blue line) as well as the overall sector average discount (green line). We can see that, typically, the JPI discount has tended to trade slightly wider (below) of the sector discount.

Source: ADS Analytics LLC, Bloomberg

We then calculate the spread or the difference between the fund and sector discount - this is simply the difference between the two lines above. We can see that the discount spread has tended to trade in the range of -4% and -2% (that is, the JPI discount was 2-4% wider of the sector discount). The other thing we notice is that, more recently, the spread has been on the wider end of the range, closer to -5% or, in other words, whereas the JPI discount has tended to trade 2-4% wider of the sector discount, it is currently trading around 5% wider of the sector discount. This means the fund's discount appears to be at the wider end of its historic range relative to the sector. To put it in terms of the fund price - the price of JPI is trading lower relative to the sector than it has tended to in the past. This may give us an indication that the fund may be cheap relative to the sector, all else equal.

The second important thing to notice is that the spread has tended to mean-revert. This gives us more confidence in making statements about the spread than if it were trending up or down through time. Nothing about the price pattern guarantees that the spread will converge back to the its historic mean but it does give us some comfort that the current price action in JPI looks to be a historic outlier.

Source: ADS Analytics LLC, Bloomberg

Finally, we convert the spread into a percentile score, ensuring there is no look-ahead i.e. the percentile of a given date only looks back into the past. As we suggested above this allows us to normalize the spread across all funds in the sector as well as quickly identify whether the current reading is compelling for a given fund (i.e. close to 0% or 100%). In our calculations, we throw out the first three months of the spread because of the technicals having to do with the fund issuance at a premium which can skew later readings as well as use a rolling 1-year window for the percentile which allows the metric to adjust to changing fund dynamics.

Source: ADS Analytics LLC, Bloomberg

Ok Fine, But Does This Actually Work?

We think there are three interesting ways to visualize whether the strategy of overweighting funds with a low discount sector spread percentile and underweighting funds with a high discount sector spread percentile makes sense.

First, going back to the chart above we can see that the discount sector spread tends to mean revert, so when the spread goes up, the fund outperforms the sector and when the spread goes down the fund underperforms the sector. This behavior runs through all funds in the sector so we are not cherry picking here.

Source: ADS Analytics LLC, Bloomberg

Secondly, we aggregate all the funds in the sector and plot the starting discount sector spread percentile against the fund's forward 3-month performance relative to the sector. The regression line below slopes in a way that shows that funds starting with a low percentile subsequently outperform the sector (the relative return is above zero) in the following three months. The total relative performance is about 2% on each side which doesn't sound like a lot but annualized over 12-months, it is 8% which is a decent amount of alpha.

Source: ADS Analytics LLC, Bloomberg

The final way we can quantify the strategy of allocating to low discount sector spread percentile funds is by borrowing from risk factor analysis. What we do is split the entire sector into 5 buckets which are rebalanced monthly and that are ranked by the percentile metric with the 0 bucket having the lowest readings. We plot the buckets total returns since 2004. It's clear that the lowest buckets (those having funds with the lowest percentile readings) outperform those with higher readings. It's interesting to see that even on a monthly basis there is a lot of alpha available.

Source: ADS Analytics LLC, Bloomberg

How Can This Go Horribly Wrong?

As the section above suggests, there appears to be real alpha in this strategy which is explained by a compelling behavioral explanation. However, it is worth keeping in mind that there are no iron laws in markets and that a strategy that works today may not work tomorrow. What are there scenarios in which the discount sector spread can send a false signal?

First, we are not volatility adjusting the discount sector spreads and so could be missing the real signal. That doesn't seem so bad to us because the signal we are using seems to work ok, most likely because funds within a given sector tend to have volatilities near each other, and volatility adjustments add an element to the strategy that may make it overly complex to execute.

Secondly, fund dynamics change - for example, a fund could begin to pursue a strategy that delivers more or less alpha than previously, which would justify a secular change in its discount spread. We partly account for this by using a 1-year rolling window so the signal could be wrong for some time but will then come back into place. The situation where this doesn't work is when a fund's strategy changes so frequently that the discount sector spread is not able to settle into a reasonable trading range.

Finally, we acknowledge that our sector determinations are not perfect and there are funds within a given sector whose strategies do not fully align. We don't have a perfect solution here. Our suggestion would be to either try to diversify this away by investing across multiple potential opportunities in the sector or do more fundamental analysis on the fund to see if there is a reason for a step change in its discount sector spread.

Where Do We Go From Here?

We think this strategy can appeal to a number of different investor types.

  • Tactical: those investors interested in gaining an edge in the market can follow the metric to pick up a few percentage points of alpha that may be available. These investors can allocate to most attractive opportunities across all sectors.
  • Core: investors with a core allocation to a particular sector can rotate their holdings towards funds within the sector that look most attractive and rebalance their positions on a 1-3m basis.
  • Systematic: investors who prefer to make portfolio allocations in a systemic way can allocate to the best opportunities across different sectors and then weight those sectors by their marginal risk e.g risk parity.
  • Pure alpha: those investors not willing to take beta exposure can potentially monetize the strategy alpha by going long the best funds and short the worst funds or alternatively, to short the benchmark ETF against the long basket of CEFs.

Going forward we will publish a table such as the one at the top of the article across a number of sectors that provides the discount sector spread to our readers.


Closed-end funds have tended to display regular patterns in how their discounts behave relative to sector averages. Discount sector spreads - or the difference between the fund's discount and the sector average has tended to mean revert with various frequencies, presenting a potential alpha-generating opportunity for fund investors. We think there is a clear behavioral explanation for this dynamic and the amount of historic alpha available in this strategy supports allocating to this strategy for various types of fund investors. We think this type of strategy is particularly compelling in a period when positive beta returns are tough to come by such as we have seen in the past year.

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.