I tend to reinvest dividends regularly in whatever equity I think represents the best value at the time. However, one idea I have frequently toyed with is to instead hold the funds in cash only reinvesting once the market (or the particular portfolio) dropped 20%. I like the idea of this second method not so much for higher returns but rather it seems logical to me that it would reduce volatility without reducing returns "that much." I continue to follow my method for now, but wonder if maybe the alternate method might be a better choice when I am less active in investments and/or less willing to take on risk.
So, I decided to do a little one-day study on some asset and see just how much of a difference it would make. I wanted something that represented a portfolio at least somewhat similar to my Search for Value portfolio 'SFV' but of course at 3.5 years, that portfolio itself did not have enough history to be useful. I wanted something that would go back at least prior to the last Great Recession the longer the better.
The S&P 500 is an obvious choice. One can get many decades of history on it. However, it is not really very representative of the Search for Value portfolio since it mainly holds large cap established stocks with relatively minor dividend payouts. The Search for Value portfolio is decidedly small cap value, with some preferreds and bonds, plus a distinct leaning towards pass-through and other securities, which pay out 8% or more in dividends. So, I needed something that more closely reflected that kind of mix.
This turned out to be much harder to find than I expected. I first did searches for ETFs. They would have been optimal because various online tools I am familiar with treat them like stocks making the rest of my job much easier. There is already software out there that will give you nice graphs with and without dividends re-invested, unfortunately, the ones I have access to do not handle mutual funds. So, I wanted an ETF.
I used various screeners but over and over again I kept running up against the same problem, most ETFs just haven't been around long enough to be useful in this study. The few that have, were not any more representative of my portfolio than the S&P 500.
Next, I looked at closed-end funds. You can find ones that are a bit more representative of my portfolio, but once again, none of the ones that were had existed long enough to be useful. Additionally, the CEFs that had been around a long time come with a significant drag from fees. When you are talking about a CEF that has averaged maybe a 5-7% return over the last 20 years, having a 1.5-3.5%+ drag from fees is very significant.
So much so that they too were really not representative of my portfolio, primarily individual equities with no fee drag. The Virtus Global Dividend and Income Fund (NYSE:ZTR) at first seemed to fit the bill with its long, 29-year history, 11% payout, and relatively reasonable 1.3% fee, but looking at the actual holdings:
It was clear this was actually a bunch of mega cap dividend payers. Reynolds American (NYSE:RAI), Altria (NYSE:MO), Shell (NYSE:RDS.A) (NYSE:RDS.B), AT&T (NYSE:T), and Vodafone (NASDAQ:VOD) might all be solid dividend payers; however, they are nothing like my small cap pass-through security centric portfolio. They are also not something I am going to pay someone 20% of the expected return to choose for me either (1.3% vs. 6.6%). Furthermore, it was clear there was no way actual dividends from the holdings of this fund were actually funding the double-digit payout, rather a combination of leverage and return of capital was being used.
ZTR was not totally useless however, I was able to quickly run an online tool, which showed ZTR returns with and without dividends reinvested:
This showed once again something I already knew and had proved to myself over and over again, dividend reinvestment matters. As you can see above, ZTR is not a particularly well-performing CEF. Its average annual return is only 2.3% since 1995, hardly worth paying a 1.3% fee to achieve. However, the 11% payout, does help emphasize the value of re-investing dividends. If you had re-invested dividends in ZTR, you would have achieved a 5.1% return. Still nothing to write home about but twice what you would do otherwise.
Put a different way, by re-investing dividends regularly, you are buying more shares when the CEF is low-priced and boosting your overall return by about 3%. I can tell you I have read a number of different academic studies that show essentially the same thing. Regular reinvestment of dividends boosts returns while also lowering volatility. It matters.
That however was not the point of this study.
Eventually, although I knew it meant a lot more effort in excel, I switched to mutual funds. Mutual funds have been around forever and there are lots of them. Surely, I could find one that put out a high amount of dividends, invested in high dividend small-cap securities as well as some preferreds and bonds, and had been around a couple of decades? Well once again, this was a much harder search than I expected, but eventually, I did find one that fit the bill: Yorktown Multi-Asset A (MUTF:APIUX). At least enough that I'm not going to continue a search that should have taken 15 minutes but instead took over 4 hours.
As you can see, Yorktown does have a sizeable holding of pass-through securities:
Ares Capital (NASDAQ:ARCC), New Residential (NYSE:NRZ), and Two Harbors (NYSE:TWO) have all been SFV holdings at one time or another; ARCC still is. Also, Apollo Commercial Real Estate (NYSE:ARI), Solar Capital (NASDAQ:SLRC) and BGC Partners (NASDAQ:BGCP) are all securities I have looked at in the past and would buy given the right price. Additionally, Yorktown keeps a meaningful amount of its portfolio in bonds and preferreds (30% currently) and has most of its holdings in US-based assets (83% currently).
This is also similar to SFV. Yorktown's long-term return has been about 7.2%, a little less than what I have done over the last 17 years, but if you add back in the high 2.5% in fees, it is much closer. Also, Yorktown throws out a meaningful 6.5% dividend (pretty close to what SFV does) and has almost 20 years of data available.
I finally found my candidate.
Yorktown was not of course exactly the same as Search for Value, but close enough. After all, I was not trying to find a good investment, rather I was just trying to find a fund somewhat similar in holdings to SFV, which had enough data for me to back-test. The other indirect benefit here is this study was definitely not curve fitted. Yorktown is a decent but not particularly notable fund, and was not chosen to prove a predetermined point.
It is just the first fund I came across which seemed appropriate. Also, it is not the absolute performance I care about anyway. It is rather the relative performance between reinvesting dividends immediately or waiting until the fund had dropped 20% to reinvest which I am interested in.
I won't waste everyone's time by posting it here, but I was able to download 19 years of daily price and dividend history for Yorktown into Excel (ask if you want a copy). I was then able to refresh my Excel skills figuring out how to write a formula which calculated whether there has been a 20% drop from the last high, keeping the dividends in cash until then, and then re-investing that cash the next day (ha, this 50-year old has still got it!).
As it turns out, this 20% drop followed by reinvestment only occurs 4 times - Sept 2008, Nov 2008, Mar 2009 (all Great Recession) and for some reason in Aug 2015. So there was a lot of cash built up between the start March 1998 and that first reinvestment in Sept 2008. We should expect quite a drag from the cash sitting there earning 0. By comparing this to the returns, which occur from reinvesting the dividends immediately, we should see just how much performance the safety of building up all that cash costs us.
As you can see, the strategy of reinvesting dividends immediately vs. holding them in cash made almost no difference in returns between inception (March 1998) and the Great Recession. $10,000 invested in the two portfolios has almost the same value whether you re-invested those dividends immediately (blue line) or let them build up cash (orange line = value in fund + cash buildup).
This was surprising to me. I had assumed there would be significant underperformance. Maybe it is because this diversified fund only returns 7% on average (I say only but that was still good enough to earn it 4 stars from Morningstar) and did have very low volatility up to the Great Recession.
Another observation was Yorktown did not have a big buildup in the late 90s only to crash in the early 2000s during the dot.com bust. This is logical as it does not typically own technology and other exciting investments but rather fixed income securities and heavy dividend payout, pass-through, equities.
Once the Great Recession hit of course things started to change. Yorktown loses over 40% from high to low during the Great Recession ($11.72 to $6.51 per share). I can tell you from other studies and personal experience that pass-through securities lost every bit as much in price as the market did during the Great Recession. They however also kept paying out pretty much the same in dividends. So, the pass-through security stock prices were affected, but the underlying businesses and cash flows those stocks represented did not take as big a hit.
All the cash that was building up in the reinvest after a 20% loss strategy got reinvested during the Great Recession. It did not get reinvested at the lows, the majority of it goes back in at a price of $9.16 a share (a 20% drop from recent highs, but still 30% higher than the eventual low at $6.51). That however does not prevent that strategy from working out well, once the market starts to recover so too do the two portfolios. One of them does has a few more shares in it than the other, the one that reinvested immediately, but it is not a large difference.
That is another surprise, that there really is not that much difference. While the reinvest after a 20% decline strategy goes from 962.5 shares before re-investment to 1401.9 post re-investment, a large change as expected, the reinvest immediately strategy already had 1,399.8 shares in it from reinvesting as each dividend was paid. I don't know if this is merely coincidence, but the difference in number of shares between the two is only 0.2% (=1,401.9/1,399.8), much less than I expected.
My last surprise observation, was that the reinvest only after a 20% decline strategy did only slightly worse than the re-invest immediately strategy. This is despite that strategy starting off by sitting there building cash for over 10 years! Ultimately, investing $10k in Yorktown in March 1998, reinvesting dividends as they got paid, resulted in an account value of $26,058 on June 16, 2017. Investing the same $10k in Yorktown on the same date but only re-investing after the fund had dropped 20% resulted in $25,662 in account value. Surprisingly, the building cash option not only had less risk, but only slightly underperformed on overall return.
Honestly, I draw no conclusions from this study. It is only one specific example over a 20-year period. I would want to see similar results from other independent sources before making a decision. However, it was an interesting start.
If we can learn one small new thing about investing every month, we can do quite well.
Disclosure: I am/we are long ARCC.
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.
Additional disclosure: This article covers a speculative investments. I do not know you: your goals, risk tolerance, or particular situation. Therefore, I cannot recommend this or for that matter any investment to you. Do your own due diligence.