Summary
- Preferred shares have outperformed both equity and treasuries with a higher return and a lower standard deviation than both in 2023.
- Both history and financial theory show that this is an anomaly.
- Over time, preferred shares are never the optimal choice to earn excess returns, and they are never the optimal choice to reduce risk.
- A simple barbell strategy with assets invested in an S&P 500 ETF and a US Treasury ETF should result in superior returns for similar or lower risk.
- Long-term investors holding preferred shares should use the current strength to re-balance their portfolios. Others should avoid them.
Three years ago, when markets were crashing and friends had seen their Rainy-Day Funds which were invested in Preferred Shares sell off 25%, I wrote an article titled Never Buy Preferred Shares - A Lesson From Baseball, The NBA, And The CDO Market. Although I walk that statement back a little, I largely stand by that advice. Retail Investors, especially, should avoid Preferred Shares as an asset class, with the potential exception of Preferred Shares issued by Utilities.
There was considerable pushback to my article. A lively debate between myself and a trader of Preferred Shares played out in the Comments Section, and several people wrote that they learned a lot from that dialogue. I encourage readers to refer to that article. It outlines lessons from the Credit Derivative and CDO markets, and it also draws analogies with the NBA and Baseball. Both sports have abandoned the "mid-range" that is their equivalent to Preferred Shares and adapted a barbell strategy. In the case of the NBA, this has manifested itself as a focus on 3-point shots, supplemented by 7-foot centers who shoot from 5-feet or less from the basket. Major League Baseball values Home Run hitters, even if they strike out a lot, or on base specialists who draw walks and hit singles. A hitter who only hits doubles contributes little to winning if there isn't already someone on base. Finally, data from the Rating Agencies was presented that shows that the "insurance", or downside protection that investors purchase through the sub-standard returns of Preferred Shares is illusory. Corporate bankruptcies happen with less frequency than meteors hitting the Earth, much less a house, and when bankruptcies do occur, Preferred Shareholders usually recover zero after lenders are paid out.
As per Table 1, the lack of downside protection isn't limited to the Preferred Shares of corporations in financial distress, it is also non-existent for the asset class as a whole when markets are under stress. So why pay up for downside side protection that isn't there?
Table 1: Returns During Market Selloffs
Time Period | SP500 | SPPREF | |
---|---|---|---|
January 2022 - September 2022 | -15.12% | -21.59% | -15.47% |
February 2020 - April 2020 | -25.37% | -21.27% | + 8.33% |
September 2018 - December 2018 | -12.10% | -10.58% | + 1.65% |
June 2015 - October 2015 | -8.90% | -3.43% | +2.05% |
April 2011 - September 2011 | -7.52% | -9.60% | +14.91% |
I. Here's What's Happened Since 2020
I analyzed the daily returns for SP500, SPPREF, S&P 7-10 Yr. US Treasury Index, and LQD , the ETF that tracks the iBoxx Investment Grade Corporate Bond Index, for the period between May 1, 2020 (the day after my original article was published) and January 31, 2023. I certainly can't be accused of cherry-picking data to support my investment thesis. On May 1, 2020, the S&P 500 sold off 2.81% whereas Preferred Shares were up 0.32%. As noted in the Bullet Points, and as per Table 2, Preferred Shares are currently enjoying a moment in the sun, with 20 positive days in January 2023 versus one negative day.
Table 2: Risk / Return Data for January 2023
SP500 | SPPREF | LQD | Intermediate Treasuries | |
Average Daily Return | 0.31% | 0.55% | 0.25% | 0.18% |
Standard Deviation | 1.05% | 0.52% | 0.64% | 0.61% |
Standard Deviation Positive Days | 0.65% | N/A* | 0.45% | 0.38% |
Standard Deviation Negative Days | 0.60% | 0.51% | 0.14% | 0.36% |
* Two negative data points are needed to calculate a Standard Deviation
I have chosen to analyze Daily Returns and Standard Deviations instead of monthly or annual figures, because many investors invest in Preferred Shares due to the mistaken belief that the asset class offers a safe harbour that can be accessed for liquidity in times of turmoil. If valuations are jumping around excessively on a daily basis, that undermines this belief. This article that was written by the CFA Institute earlier this week is on point. It discusses the historic daily and monthly correlation of various asset classes, and the differences between them.
It is also worth noting the difference between the classic definition of risk in Financial Theory (which is Standard Deviation from the mean) and Downside Protection. Financial theory states that if there are two stocks, A and B, where A decreases 5% a day and then increases 4% a day, and B also decreases 5% a day but then increases 10% a day, B is the riskier stock. This counterintuitive conclusion is because the volatility of B's returns is greater even though most of the variance is to the upside. Sharpe Ratios are based upon this philosophy, whereas Sortino Ratios only consider negative returns and harmful volatility to the downside. It is for this reason that I show the Standard Deviation for both positive and negative days. As per Table 3, much of the so-called "excess risk" of Common Equity is due to its greater positive upside variance from the average return. The Standard Deviation of Negative Days is much closer to that of other asset classes.
Preferred Shares have underperformed over the long-term, (as per my previous article) and, as per Table 3, they have also underperformed compared to equities over the past 34 months. The outperformance of the S&P 500 versus SPPREF of 6 bps per day may seem trivial, but it adds up. For the period as a whole, the S&P 500 Index has increased by 44%, whereas SPPREF has fallen by 4.75%. This underperformance has also been the case when the two asset classes are compared on an annual basis, in 2020, 2021 and 2022. It is only the month of January 2023 that Preferred Shares have outperformed.
We have just experienced the highest inflation and worst bond markets for 50 years. Despite this, the average daily return for Treasuries versus Preferred Shares is a push, both bonds and preferred shares have had a slightly negative average daily return, -0.01% and -0.0048%. However, what isn't comparable are the risk figures. Preferred shares has a Daily Standard Deviation that is 28% greater than the Daily Standard of Intermediate Treasuries. On Negative Days, this figure rises to 42%. So, the optimal instrument for intermediate and long term positive returns is Common Equity. The Optimal instrument for reducing risk is US Treasuries.
Table 3: Risk / Return Data for May 1, 2020 - January 31, 2023
SP500 | SPPREF | LQD | Intermediate Treasuries | |
Average Daily Return | 0.06% | -.0048% | -0.02% | 0.01% |
Standard Deviation | 1.22% | 0.59% | 0.57% | 0.46-% |
Standard Deviation Positive Days | 0.74% | 0.41% | 0.38% | 0.32% |
Standard Deviation Negative Days | 0.88% | 0.44% | 0.39% | 0.31% |
II. Why Preferred Share Consistently Underperform on a Risk / Return Basis
Preferred Shares suffer from two fundamental asymmetries. The first is asymmetric returns. Preferred Shares and Common Shares both have the same maximum downside - 100%. However, the upside for equity is theoretically unlimited, whereas a Preferred Share's upside is capped by its call price. So, positive days are greater in scale for equity, and these gains offset the losses on negative days, in a way that Preferred shares don't. It is worth noting that the same asymmetry is present in Corporate Bonds, but whereas Preferred Shares rarely pay out anything in the unlikely event of a corporate bankruptcy, bonds and loans, depending upon their seniority, usually do.
The second asymmetry concerns the incentives and information between issuers of Preferred shares, and certain classes of investors that act as buyers. Corporations have a fiduciary duty to act in the best interests of their owners, and managers are paid in part with common equity to align their interests with shareholders. This fiduciary duty and alignment of interests does not extend to Preferred Shareholders. Debt makes up a large portion of a typical corporation's Balance Sheet, and lenders who are usually sophisticated investors, are able to use this negotiating power for their own advantage. In short, they will not agree to terms that hurt their own interests for the benefit of Preferred Shareholders. If a corporation goes to the trouble of issuing Preferred Shares, and if lenders allow this, it is because they both believe that they benefit from this issuance.
To illustrate this, consider the following hypothetical example of a corporation that owns $100 million of Assets (the left-hand side of the Balance Sheet), financed equally with Debt and Equity (the right-hand side of the Balance Sheet).
Assets = Debt + Equity
$100 million $50 million $50 million
Now assume that this issuer goes out and buys a further $10 million of identical assets - maybe it's an mREIT buying more of an MBS that it already owns. Further assume that it finances this purchase with $10 million of Preferred Shares. The new Balance Sheet is as follows:
Assets = Debt + Preferred Shares + Equity
$110 million $50 million million million
If anything, the debt is now worth more than $50 million. It has $110 million of assets as coverage instead of $100 million - its Credit Rating might even be upgraded. That leaves a maximum amount of $60 million to be split between the Equity holders (who started at $50 million) and the Preferred Shareholders. Yes, I know, synergies can make the assets more valuable, I get it, that $10 million of new MBS is going to make that $100 million of identical old MBS more valuable. Or maybe that $10 million plot of land will make the $100 million of real estate already owned more valuable . Actually, this latter scenario is possible, but why would equity holders leave value on the table for Preferred Shareholders, when they can keep it all for themselves by issuing new debt?
While this is a simplistic example, it's worth examining the concept of an Arbitrage CDO. Like banks, they own financial assets, Loans and Bonds, or in the case of mREITs, MBS. Like banks, they fund themselves with various classes of debt and the equivalent of Preferred Shares - the second and third loss tranches of a CDO. The value of their assets is set by the market - there are no synergies to be had. Arbitrage CDOs encash the credit spread differential between the high-yield securities that form the collateral (the rate of return on the assets acquired by the CDO) and the low-yield liabilities represented by the rated CDO securities (the funding cost of the liabilities issued by the CDO). In short, they don't arbitrage the financial assets on the left-hand side of the Balance Sheet, they arbitrage the liabilities and financial instruments on the right-hand side of the Balance Sheet. That would be Debt Holders or Preferred Shareholders. As they say in poker, if you don't know who the sucker at the table is, look in the mirror.
III. Are there any Preferred Shares that, in general, are worth considering?
Banks have no practical limitations on the assets that they can originate. In many ways, banks are similar to Arbitrage CDOs or mREITs. All three types of entities seek to profit by clipping the spread between the return they earn from the financial assets that they own, and the lower yield they pay for the financial liabilities that they owe. In the extreme, if a bank can raise enough cheap capital, it can buy another bank. So there is no natural cap on the amount of shares that they will issue.
Regulated entities like utilities are different than banks. They can't get out a cookie cutter and build a new Hydro Electric Dam worth $25 Billion every other month, so there is a cap on the supply of Preferred Shares from these types of issuers. Further, they are tightly regulated. Regulators do not want their shareholders earning fat returns for a low-risk monopoly business, because it means that consumers are getting over charged. Rates are set using a cost of capital formula, where certain assumptions are made regarding leverage, the cost of debt, and an allowable fair return for common equity holders. Once these agreements are in place, it may make sense from time to time for a utility to Arbitrage these rules. This is an important distinction - it isn't investors being taken advantage of. The ultimate source of the value created is consumers who are paying more than they otherwise would have. In such cases, value is created by issuing a small tranche of Pref shares that improve the quality of debt, and this value can be split between debt holders, equity holders, and the new preferred shareholders.
PFF is the largest ETF of Preferred Shares. It has $13 Billion of assets, 67% of which are invested in the Preferred Shares of Financial Institutions. PFXF , the VanEck Preferred ex Financials ETF ($1 Billion of AUM) has no exposures to Financial Institutions, and it has a slightly lower expense ratio than PFF, 40 bps versus 46 bps. The two indices are highly correlated, although PGX has slightly outperformed PFF while exhibiting less volatility. When compared to SPY however, all that you can really say is that it has been the best of a bad bunch.
IV. Okay - you may have a point, but what are the alternatives?
People who are a lot smarter than me have said that the only free lunch in the financial markets is diversification. Usually this refers to a well-diversified portfolio of 30 or more equities. However, diversifying across different asset classes can also provide benefits. Table 4 shows the performance of a simple barbell strategy of the SP500 and Intermediate Treasuries versus SPPREF. The weighting of the two asset classes (59% Bonds and 41% SP500) has been adjusted so that the Daily Standard Deviation of the portfolio equates to that of SPPREF. As can be seen, the two-asset portfolio has outperformed SPPREF by 2 bps per day, or 16% over the entire 34 month period, with no additional risk (as measured by Standard Deviation) being incurred. When considering downside protection, the Maximum Daily Loss and Standard Deviation on Negative Days is lower for the portfolio. Preferred Shares also offer some diversification benefits, but not to the same extent as bonds do. Being a hybrid, Preferred Shares are more closely correlated with Common Equities than Bonds are, and as already seen in Table 1, they do not exhibit the same counter cyclicality.
Table 4: Risk / Return Data - SPPREF vs. Portfolio of 36% Intermediate Treasuries and 64% S&P 500, May 1, 2020 - January 31, 2023
SPPREF | Portfolio | |
Average Daily Return | -0.01% | 0.01% |
Maximum Daily Gain | 4.28% | 3.60% |
Maximum Daily Loss | -3.25% | -2.63% |
Standard Deviation | 0.59% | 0.59% |
Standard Deviation Positive Days | 0.41% | 0.37% |
Standard Deviation Negative Days | 0.44% | 0.43% |
V. Conclusion
History has shown that Preferred Shares do not provide what the label on the can says they do. They are hybrid instruments which do not offer the safety of debt in a bankruptcy, or the stability of bonds in volatile markets. As their upside is capped, positive returns do not offset periods when there are negative returns to the same extent that common equity does. Further, there are information asymmetries and competing incentives for issuers versus investors. The result is that they are often sold at prices which do not compensate investors for the risks being incurred.
Diversification across different classes will offer superior returns for the same risk; long term investors should view this period of strength as an opportunity to re-balance their portfolio. As always, consult advisors who are familiar with your unique circumstances and risk tolerance.
For further details see:
Preferred Shares Are Outperforming, It's A Great Time To Exit PFF And The Asset Class