Conclusion: CEFs that have current premiums/discounts nominally lower than their four quarterly rolling averages have in general experienced greater stock price performance in post 3 and 6 month periods than those CEFs with premiums/discounts nominally greater than their comparable rolling average. Nominally lower means that the discount would be greater (-15% versus -10%) and premiums would be less (10% versus 15%).
This was particularly true for the “Bottom 3” versus the “Top 3” CEFs for a post 6 month period (see graph below).
Observations: The adjacent chart represents the highest and lowest CEF current premium/discount spreads as of the quarter-end June ’10 compared to their four quarter-end rolling average (June, Mar, Dec and Sep).
Based on historical results sighted below, the CEFs with the bottom-half spreads could outperform those of the top-half spreads for a 6 month period ending 12/31/10. In particular this should be true of the bottom 3 (“green”) and top 3 (“yellow”).
Background: The old “saw” for closed-end fund (CEF) investors is to: “Buy a CEF’s stock at a discount and sell it when it goes to a premium”. Unfortunately, on a nominal basis this is rarely the case.
Most CEFs don’t gravitate to par as a long-term average. In fact, the industry historically trades at discount and individual CEF’s long-term premium or discount can vary widely from par.
So, it would seem intuitive that CEFs that trade significantly from their average long-term premium or discount would have a tendency to gravitate towards their long-term average. This seems to be the case.
Methodology: A CEF’s quarter-end premium/discount was compared with the rolling average of its past four quarters’ premium/discount to generate a spread between the two results (“Cur-RollAvgSprd”).
The standard deviation was calculated for each quarter-end for 580 CEFs based on their Cur-RollAvgSprd for each quarter from 12/08 to 6/10. Those CEFs that registered a premium/discount that was greater or less than 2 standard deviations from the average were accumulated into a group for each quarter-end (“SD2CEFGroup”).
Stock price changes for the SD2CEFGroup were calculated for post 3 and 6 month periods. The average stock price percentage change for the top-half of the SD2CEFGroup (high spreads) was compared to the bottom-half of the CEFs with the low spreads.
Results: There was a tendency for those CEFs with lower spreads (bottom-half) on average to outperform those with higher spreads (top-half) on a 3 and 6 month bases for the December ’08 to December ’09 period. This was particularly true for the Bottom and Top 3 CEFs.Click to enlarge
Caveats: There are many factors that impact a CEF price performance. It seems the most important are: nominal yield, sector momentum and relative premium/discount.
The current article focused on the current premium/discount less the four quarter rolling average (“Cur-RollAvgSprd”) versus future price changes.
A different rolling average (monthly) and different post price change periods could have produced different results. The period under study was limited to 12/08 to 12/09 and a longer interval may have also delivered different results. CEF distributions were not considered for measurement periods.
Additionally, a CEF with a widening discount spread can be interpreted negatively as it may be indicative of a CEF having a less attractive attributes e.g., higher leverage, an impending negative event, i.e., distribution cut, or subject to sector rotation.
The article made no distinction between widening discount and narrowing premiums that on further study could provide additional insight.
Disclosure: Joe is both long and “short” a diversified portfolio of CEFs.