The purpose of this article is to tell you about the dividend/EPS ratio and its predictive properties when it comes to total return expectations and current valuations. The story will be told with spreadsheets. I will end by touching on the use of that ratio in explaining current BDC (business development company) valuations, using Solar Capital (NASDAQ:SLRC) and Triangle Capital (NYSE:TCAP) as examples. The stats used are for every dividend-paying BDC in my coverage universe.
We live in a yield plus CAGR (compound annual dividend growth rate) world -- with some adjustments for risk -- when it comes to stock valuations and total returns. It is my goal as an individual stock investor to understand and use those valuations. You are not going to win as an investor by buying overvalued stocks. At the same time, many stocks with above sector average valuations are still undervalued. I want to use a nuanced view of valuations. It is my goal to maximize my total returns while accepting a lower amount of risk along the way. There are stocks that are the canaries in the coal mine. I want to minimize my ownership of the canaries. You are not going to win as an investor by owning dead birds.
In all sectors where I invest, I use the dividend/earnings ratio as one of the most important tools in assessing dividend safety and dividend growth. Most often this is the dividend/EPS ratio. With REITs, it's the dividend/AFFO ratio. With MLPs, it's the distribution/DCF (distributable cash flow per unit) ratio. With BDCs, it is the dividend/NII (net investment income) ratio. Due to BDC EPS projections (which one can find at Yahoo Finance, Google Finance, CNN Money, etc.) being NII projections, we can also call it the dividend/EPS ratio for BDCs. The P/E ratio and price/NII ratio is the same thing.
I have owned banks and REITs that have cut their dividends. If you have been a dividend investor since 2007, chances are you have had that experience, too. The loss in share price that accompanies dividends cuts -- or even the anticipation of a cut -- is unpleasant. We all want to avoid that. At the same time, I have owned one or two MLPs that have underperformed in distribution growth. Those lower growth MLPs have underperformed in unit price appreciation. I have not let ownership inertia or the unpleasant tax consequences keep me from cutting loose some of that dead weight. Any tool that provides a picture of dividend growth and dividend security is one worth having and using.
To show that this metric is meaningful when it comes to total returns, we should look at a few years of historical data. If the information in this ratio is already fully priced in to the stocks, then it will not assist us in generating better total returns. I believe the data will show that the information is only partially priced in to the current valuations.
2012 Spreadsheets -- BDCs Dec. 31, 2012
Yields in the spreadsheet below are based on the Q4 2012 dividends. KED, NGPC, PSEC, and TAXI have announced increases in their Q4 2012 dividends. The Q3 2012 NAV (net asset value per share) is used in the price/NAV. ARCC paid a special $0.05/share dividend in Q3 and Q4, which is not included in the stats below. I am using the Q3 2012 DNOI (distributable net operating income) run rate as the KED and TTO 2012 EPS estimate.
Roll over for larger tables.
|Share Price||12 EPS Est.||Div/||Div/||Q3-11||Q3-12||Price||YTD Percent Change|
|With the 10 Treasury at 1.76% and sector average yield [on Q4 Divs] at 9.37% - the spread is 761 bps.|
2012 Stats for BDCs With Q1 2012 Div/EPS Ratios Under 95%
|Share Price||12 EPS Est.||Div/||Div/||Q3-11||Q3-12||Price||YTD Percent Change||LTM Change|
2012 Stats for BDCs With Q1 2012 Div/EPS Ratios Over 95%
|Share Price||12 EPS Est.||Div/||Div/||Q3-11||Q3-12||Price||YTD Percent Change||LTM Change|
For dividend paying BDCs of 2012, those with beginning of the year EPS projections and the Q1 2012 dividends that combined to make a dividend/EPS ratio of less than 95% had a mean price gain for the year was 19.20%. Their mean dividend growth for the year was 17.35%. Their mean total return for the year was 29.70%. And seven of the 14 beat the sector median yearly price gain. Of these 14, every one but GLAD and GAIN had dividend growth. Both GAIN and GLAD have 2013 EPS projections below their 2011 EPS actual. The 17.35% mean growth was upwardly skewed by large dividend gains by KCAP, MCC, and SUNS.
The following companies had dividend/EPS ratios of greater than 95%: AINV, BKCC, FSC, FULL, GBDC, HRZN, MAIN, MCGC, NMFC, NGPC, PSEC, SLRC, and TICC. Their mean price gain for the year was 13.5%. Their mean dividend growth for the year was -3.49%. Their mean total return for the year was 24.8%. And three of the 13 beat the sector median yearly price gain. Of these 13, MAIN and NMFC had double-digit dividend growth, and PSEC had over 8% dividend growth. Both MAIN and PSEC started the year with very inaccurate and low EPS projections. NMFC had some growth in its 2012 EPS projections during the year. The 5.49% mean negative dividend growth was downwardly skewed by large dividend cuts by AINV, FSC, HRZN, and NGPC.
The (28.7% - 24.8%) 3.9 percentage point difference in total returns for these two groups is small. But the average returns covers up the story. AINV began the year undervalued. It is my belief that the market knew of the then upcoming dividend cut -- but failed to correctly size that cut. AINV began the year with a price/NAV 11 percentage points below average, a yield 7.71 percentage points above average, and a P/E 1.92 points below average. Those are the metrics of a bad BDC. AINV is closer to average than it is to being bad. AINV spent 2012 catching up to something close to sector average on valuations like yields, price to NAV and P/Es. MAIN, which did not belong in the low-dividend-to-EPS-ratio group, positively skewed their average a good deal.
Despite the number of exceptions that created noise in the data, it should be clear that the dividend to EPS ratio strongly influenced the resulting dividend growth. And dividend growth matters when it comes to total returns. The following companies had dividends that grew more than 1% between Q4 2011 and Q4 2012: ARCC, FDUS, HTGC, KED, KCAP, MAIN, MCC, NMFC, PSEC, PFLT, PNNT, SUNS, TAXI, TCAP, TCRD, and TICC. Their mean price gain for the year was 22.18%. Their mean total return for the year was 33.04%. And eight of the 16 beat the sector median yearly price gain.
The following companies had dividends that grew less than 1%: AINV, BKCC, FSC, FULL, GAIN, GBDC, GLAD, HRZN, MCGC, NGPC, and SLRC. Their mean price gain for the year was 8.13%. Their mean total return for the year was 19.05% -- and two of the 11 beat the sector median yearly price gain. None of the group of beginning of the year Dividend/EPS ratio under 95% BDCs had a dividend cut. The average dividend growth for this group was 17.35%. Five of the 13 in the over 95% ratio BDCs had a dividend cut. Even with the inclusion of MAIN and PSEC, their average dividend change was negative.
Now let's repeat the process done above for the 2011 data.
BDCs -- Dec. 30, 2011
Note: Yields based on the Q1 2012 dividends -- and dividend/EPS ratios are based on the 2012 EPS estimates in the data in the spreadsheets below. ARCC and PNNT raised their Q1 2012 dividend by a penny. FSC has lowered its Q1 2012 dividend from $0.1066 per month to $0.0958. Three of the four BDCs with double-digit dividend increases -- KED, MCGC and TAXI -- had dividend cuts in prior years. SAR's $3.00 Q4 2011 dividend is ignored in this data and is given a zero dividend expectation for Q1 2012. PNNT has raised its dividend twice in the last four quarters and TCAP has raised its dividend three out of the last four quarters -- both of which are probably not a good run rates.
|Share Price||11 EPS Est.||Div/||Div/||Q3-10||Q3-11||Price||YTD Percent Change|
|With the 10 Treasury at 1.874% and sector average yield [on Q1 Dividends] at 10.6% - the spread is 873 bps.|
*DIV - in the last column, the Dividend change is calculated as the percentage difference between the Q3-11 and the Q3-10 dividend.
|For this spreadsheet, both the average yield and the average Div to EPS ratio were divided by 20 - and not 21 - to filter out the effect of the zero dividend paying ACAS.|
Dividend dates used were:
|ACAS : none||AINV : 10-03-11||ARCC: 9-29-11||BKCC : 1-04-12||FSC : monthly|
|GAIN : monthly||GLAD : monthly||HTGC: 11-29-11||KCAP : 01-27-12||KED : 10-19-11|
|MAIN: monthly||MCGC: 01-13-12||NGPC: 01-06-12||PNNT : 01-03-12||PSEC : monthly|
|SAR : 12-30-11||SLRC: 10-04-11||TAXI : 09-02-11||TCAP : 12-28-11||TICC : 12-30-11||TTO : 11-30-11|
2011 Stats for BDCs With Q1 2011 Dividend/EPS Ratios Under 95%
|Share Price||11 EPS Est.||Div/||Div/||Q3-10||Q3-11||Price||YTD Percent Change||LTM Change|
2011 Stats for BDCs With Q1 2011 Dividend/EPS Ratios Over 95%
|Share Price||11 EPS Est.||Div/||Div/||Q3-10||Q3-11||Price||YTD Percent Change||LTM Change|
For dividend-paying BDCs of 2011, the following companies had dividend/EPS ratios -- using the beginning of the year EPS projections and the Q1 2012 dividend -- of less than 95% at the beginning of 2012: ARCC, GAIN, GLAD, HTGC, KED, NGPC, PNNT, TAXI, and TTO. Their mean price gain for the year is -2.81%. Their mean dividend growth for the year is 11.27%. Their mean total return for the year is 5.19%. And six of the nine beat the sector median yearly price gain. Six of this nine had dividend growth, Three had no dividend change.
The following companies had dividend/EPS ratios of greater than 95%: AINV, BKCC, FSC, KCAP, MAIN, PSEC, SLRC, TCAP, and TICC. Their mean price gain for the year is -14.33%. Their mean dividend growth for the year is 1.3%. Their mean total return for the year is -4.44%. And four of the nine beat the sector median yearly price gain. Of these nine, KCAP, MAIN, TCAP, and TICC had dividend growth. KCAP, MAIN, and TCAP all had EPS projection growth -- and all three had total returns above the sector average.
Now let's repeat the process done above for the 2010 data.
BDCs -- Dec. 31, 2010
Note: FSC added June 16, 2010. Dividends have been in flux. Yields based on the Q4 2010 dividends are shown in this data. PNNT has increased its Q1 2010 and Q2 2010 dividend. TICC increased its Q2, Q3, and Q4 dividend. KCAP, TAXI, and TTO decreased their Q2 dividends. MCGC restarted paying dividends in July. MCGC increased its dividend to $0.12 for Q4 2010. SAR's 10-to-1 reverse stock split and its $4.40 dividend that was 90% paid in stock are not shown in stats below.
|Share Price||10 EPS Est.||Div/||Q3-09||Q3-10||Price||YTD Percent Change|
|With the 10 Treasury at 3.2877% and the sector average yield [on Q4 Divs] at 8.38% - the spread is 509 basis points.|
The dividend/EPS ratio is a logical omen of future dividend growth -- and that tends to lead to better total returns. But the dividend/EPS ratio failed to be predictive of investment success for 2010 due to there being so many failures in the accuracy of the EPS projections. Of the 20 projections, 12 had errors over 10%. Without accurate EPS projections, the ratio is meaningless. The good news: Compared to 2010, a growing number of major brokerages now cover BDCs. And the inclusion of the major brokerages has significantly improved the accuracy of the consensus projections. The bad news: In a sector where the current 2013 dividend/EPS ratio is 96.4%, an error in the EPS projection of only 5% is significant when it comes to projecting dividend growth.
For dividend paying BDCs of 2010, the following companies had dividend/EPS ratios -- using the beginning of the year EPS projections and the Q1 2010 dividend -- of less than 95% at the beginning of 2010: AINV, FSC, HTGC, NGPC, and PNNT. Their mean price gain for the year was 15.86%. Their mean dividend growth for the year was 1.37%. Their mean total return for the year was 25.54%. And 0 of the 5 beat the sector median yearly price gain.
Of these five, four had over 10% decreases in their EPS projections. Only two had dividend increases. The following companies had dividend/EPS ratios of greater than 95%: ARCC, BKCC, GAIN, GLAD, KED, KCAP, MAIN, PSEC, TAXI, TCAP, TICC, and TTO. Their mean price gain for the year was 34.43%. Their mean dividend growth for the year was -4.21%. Their mean total return for the year was 45.55%. And five of the 12 beat the sector median yearly price gain.
In summation for the historical data, 2010 was a terrible year for the predictive ability of the dividend to EPS ratio. But the cause was due to a large number of prediction inaccuracies in the consensus analyst EPS projections. The consensus projections have grown more accurate over time -- but they are still not perfect. This is why I do NII projections for almost all of the BDCs that I update and all the BDCs that I own. If I cannot have a decent degree of faith in an EPS projection, then I have no idea how to value the BDC. If I have no idea how to value the BDC, then I have no business buying it.
I do not use the same degree of discipline in all the sectors where I invest. But you can potentially loose a lot of money quickly with BDCs. EPS projections are volatile and have a history of being less than accurate. The companies in which I invest do not have multiple decades of successful total return history. The difference between a good dividend/EPS ratio and a bad one is a few percentage points. In a sector like consumer staples, I have 40 years to great dividend and EPS growth history, which generates a much higher level of confidence and trust. The dividend to EPS ratios are frequently under 45% -- so the stocks and their dividends can weather EPS surprises. The difference between a good and bad dividend/EPS ratio is huge. EPS projections are not volatile -- and the consensus projections have a history of being accurate. This is a sector where I can trust the analysts, but that may be due to it being a sector where the analysts have an easier job. This is a sector where I can budget a significantly lower amount of maintenance due diligence.
We cannot leave the topic of the EPS/dividend ratio without looking at the current data and showing how that metric is meaningful when it comes to valuations. Dividend to EPS ratios (for dividend-paying BDCs) range from as low as 78% to 114% (at April 12, 2013) and may look all too random. But this ratio helps explain the variety of yields, price-to-book ratios, and price-to-NII ratios. There is some correlation to YTD price changes as well. This first test is done for BDCs with weighted average portfolio yields of more than 10%. Lower risk BDCs have their metrics shown separately.
The following dividend paying BDCs had Q1 2013 dividend/EPS ratios of less than 93%: ARCC, FDUS, HTGC, KCAP, MAIN, PSEC, TCAP, and TCRD. Their YTD mean price gain is 5.02% and four of the eight beat the sector mean yearly price gain (3.83%). Their mean yield is 8.68% and they sold at an average price/book ratio of 1.30 and an average price/NII ratio of 10.84.
Any sub-sector that includes MAIN and TCAP is going to have a slightly skewed average for its price/NAV ratio. The price-to-NAV ratio for the subset of BDCs is 1.30 compared to the sector average of 1.09. How should one use this tool of looking at a context sensitive NAV average? Look at TCRD and its 8.97% yield, which is just slightly below the sector average. And look at its price to NAV at 1.11, which is also just slightly above sector average. Based on a comparison to sector average, TCRD looks pricey. But based on a context sensitive sub-sector averages, it is cheap. This relative valuation makes TCRD worthy of further investigation. But that task will be left for another article. I strongly believe that BDCs with good CAGRs sell at too large a discount. I believe that a winning BDC portfolio needs to be heavily weighted in the above stocks.
The following BDCs had Q1 2013 dividend/EPS ratios of more than 94% -- but less than 100% AINV, GAIN, GBDC, GLAD, HRZN, MCC, NMFC, and SLRC . Their YTD mean price gain is 1.49% and three of the eight beat the sector mean yearly price gain. Their mean yield is 9.24%. They sold at an average price/book ratio of1.03 and an average price/NII ratio of 10.68. Compared to BDCs with great dividend/EPS ratios, the good ratios are relatively cheap. I do not mind having a few "good" BDCs if they are purchased at attractive yields. And SLRC with its over 10% yields and recent entry into "covered divided" territory makes it worthy of further investigation.
The following companies had Q1 2013 dividend/EPS ratios of more than 100%: BKCC, FSC, FULL, MCGC, and PNNT. Their YTD mean price gain is 1.44% and two of the five beat the sector mean yearly price gain. Their mean yield is 10.81%. They sold at an average price/book ratio of 1.01 and an average price/NII ratio of 9.84. These are the relatively bad BDCs. There needs to be errors in their EPS projections before any of the above BDCs should be in your portfolio. I currently believe that the projections for PNNT are too low and thus in error.
The BDCs that had weighted average portfolio yields of less than 10% (the second test) are KED, PFLT, SUNS, and TAXI. Their YTD mean price gain = 8.91% and three of the four beat the sector mean yearly price gain. Their mean yield is 6.98%. They sold at an average price/book ratio of 1.13 and an average price/NII ratio of 12.78. These are the safer BDCs. And you want your BDC portfolio to be safe. The heavier your weighting in BDCs, the more of these safe and lower-yielding BDCs need to be components of your portfolio.
Let's end by answering the questions in the headline. Why does relatively safe SLRC (its portfolio is close to 50% in secured loans) yield 10.35%, while the relatively risky TCAP (with a portfolio that is 79% in subordinated debt) yield 7.83%? By now you know the answer. Look at the dividend/EPS ratio. With a dividend/EPS of 99.6%, SLRC is probably going to have another year of zero dividend growth. If dividend growth does happen, it will probably be small. I will give it a yield plus CAGR of 11% just to cover that potential upside. With a dividend/EPS of 93.5%, TCAP's ratio indicates another year of mid-single-digit dividend growth, giving it a yield plus CAGR of over 12%.