Reaves Portfolio 1 Performance During Periods of Rising Interest Rates

Transcription

1 Reaves Portfolio 1 Performance During Periods of Rising Interest Rates February 2014 Overview Over the past nine months since the Federal Reserve first publicly discussed the timing of tapering of quantitative easing, we have encountered investor concerns about the potentially adverse impact the unwinding of Fed policy might have on our portfolios. The key concern is that rising interest rates will automatically produce negative results for our type of investments, an assertion which is not supported by the historical performance data. A secondary issue is that a move higher in rates could potentially be quite sudden and severe given the unparalleled easing that has occurred since the onset of the financial crisis in Another related concern is that investors with above average allocations to dividendpaying equities in response to Fed policies will aggressively reduce these holdings and drive the market prices of interest rate sensitive stocks lower. While we share the general uneasiness of investors regarding the impact of rising rates on equity prices, we take comfort in the fact that our portfolios have performed well during prior periods of rate increases. The data presented here shows that the Reaves ERISA Composite (the ERISA Composite) generated a positive return in six of seven periods when 10-year Treasury yields rose over the past 31 years. The average gain of the ERISA Composite over these periods was 24.6% which outdistanced the Citi 10-year Treasury Index average negative return of -4.8% and trailed the S&P return of 30.4% for the same periods. The ERISA Composite produced a positive return in each of the six identified periods of Fed tightening since Its average return over the six periods was 29.0% compared to 17.2% for the S&P 500. These results may come as a surprise to the casual observer but we think they are the outcome of a carefully constructed, actively managed portfolio designed to take advantage of a rising dividend stream from companies with relatively predictable earnings growth. In addition, exposure to the energy sector acts as a natural inflation hedge. 1

2 Some investors are concerned that given sustained easing by the Federal Reserve since the onset of the global economic crisis in 2008, the expected rise in rates this time around could be quite sharp and perhaps historically unprecedented. While short-term rates could rise significantly from their current levels, we think that a review of TIPS 3 pricing and the historical average slope of the yield curve suggests that 10 to 30-year Treasuries should yield between 3% and 4%. That is to say that the market may have already discounted much of the anticipated rise in rates. As to the worry about a mass exodus from dividend-paying stocks, we think that for any period but the shortest time frame our emphasis on companies that provide a significant portion of the operating infrastructure of the U.S. economy, with their attendant capability to derive growing earnings from typically non-discretionary products and services, insulate the portfolio from short-term changes in supply/demand factors affecting any particular security. The relationship between interest rates generally and stock price performance specifically has been studied and debated throughout modern financial times. Currently, this topic is of particular interest to investors following an extended period of monetary easing and given the anticipation of a rise in rates as this easing is reduced or eliminated. In this paper, we consider various historical periods of rising rates, along with an analysis of the circumstances surrounding the data associated with those periods in an attempt to imagine what the impact of rising interest rates and/or future Fed actions might be on the stocks and sectors in which we invest. Considering History A review of interest rates, as represented by the generic 10-year U.S. Treasury yield, reveals that three years subsequent to the inception of our portfolio management activities in 1978 interest rates entered a secular downtrend. This downtrend can be traced to the relative peak of rates in 1981 to the most recent lows in Over the course of our 30+-year history there have been several significant reversals in this long-term trend. With the help of Bloomberg we have identified this long-term trajectory, along with the most significant visual reversals, in the chart that follows: Source: Bloomberg 2

3 The periods that have been highlighted in the chart share one common theme; that being of a radical and/or significant period of reversal in rates, followed (with the exception - as yet - of the most recent observation) with a resumption of the previous downtrend. We have identified the starting and ending points of these movements, along with their attendant metrics in the following table: Generic 10-Year U.S. Treasury: Historical Periods of Rising Rates Onset Yield Peak Yield Duration (Months) Absolute Move Yield Change % Dec % Jun % % 33.22% Dec % Sep % % 32.72% Sep % Dec % % 45.32% Sep % Dec % % 45.64% Jun % Jun % % 46.15% Dec % Dec % % 73.33% Sep % Dec % % 85.32% Average % 51.67% Source: Bloomberg Note that the average term of these moves has been slightly more than 17 months, with the longest period being from June of 2003 through June of 2006 and the shortest from December of 1986 through September of We list the movement from trough to peak in both absolute terms as well as the total movement as a percentage of the baseline yield. It is interesting to note that while the most recent move is about two-thirds the average absolute level (at 140 basis points) it is the highest in terms of relative change (at 85%). We would also note that if September of 2012 was the start of the move and if it conformed to the average life of past moves this would imply that the peak would occur in roughly February of As a first point of departure, we thought it might be informative to consider the performance of related indices, along with the ERISA Composite, over each of the periods as set forth in the previous table. The cumulative performance over the corresponding period of interest rate increase is as follows: Relative Performance During Periods of Rising Rates Onset Yield Peak Yield Fixed Income Reaves ERISA Composite S&P 500 Dow Jones Utilities Dec % Jun % 1.33% 40.59% 18.79% 23.20% Dec % Sep % -9.36% 4.10% 32.38% -3.73% Sep % Dec % -8.49% -9.41% 2.88% % Sep % Dec % -4.60% 22.69% 56.23% 6.83% Jun % Jun % 0.06% 72.56% 39.32% 97.01% Dec % Dec % -4.17% 22.04% 27.81% 9.37% Sep % Dec % -8.48% 19.75% 35.08% 10.43% Average -4.82% 24.62% 30.36% 17.20% Notes on Chart Data: Fixed income is represented by the Citi 10-year U.S. Treasury Index, which measures the performance of 10-year U.S. Treasury securities; Reaves ERISA Composite (Net of fees); Utilities are represented by the Dow Jones Utility Index, with pre-1984 quarterly returns converted to monthly compound equivalents. The S&P 500 Index is a cap-weighted representation of U.S. large-cap stocks. 3

4 As might be expected, the fixed income sector produced a negative return on average. Perhaps surprisingly, utilities were positive in five of the seven periods identified, whereas our ERISA Composite produced a negative return in only one out of seven periods: September 1993 to December The broad equity market produced an average 30% return while the ERISA Composite comes in a close second with an average return of 25%. We believe this is impressive given that since inception the ERISA Composite has exhibited only 86% of the volatility of the S&P 500 with an effective beta of less than 0.6. A quick visual inspection of the performance does not reveal any particularly obvious correlations between duration (term of the fixed income drawdown) or severity (either absolute or relative rate rise) and subsequent performance, but does contradict common perception that when interest rates rise stocks (value oriented/high dividend stocks in particular) do poorly. Federal Reserve and Interest Rates As a next step we thought it might be useful to attempt to differentiate among the primary drivers of past increases in rates. While there are never simple answers to matters of economic and market complexity, in broad terms increases in interest rates are generally considered to be associated with three primary (and typically inter-related) scenarios: An increase in expected or realized inflation Federal Reserve (or to a lesser extent Fiscal) activity Short and long-term impacts on supply and/or demand Changes in interest rates due to changes in aggregate demand or supply are typically more gradual, with their impact more likely realized at the margin. These changes often involve scarcity (due to a demographically driven need for yield, for example) or excess (the current trajectory of government borrowing, for example). The single and most notable exception to this rule is a sudden and coordinated rush to safe haven assets following some unanticipated and impactful event, which although temporary, has the ability to move overall rates substantially and quickly. These movements are most typically in a downward, not upward trajectory, and, as such, do not factor into this analysis. More often increases in interest rates are driven by the market s expectations for the future where nominal rates are subject to changes in realized or anticipated inflation. These changes occur simultaneous to the impact of Fed policy, which is typically attempting to react to (and manage) the same underlying economic variables. Absent sudden price shocks to the system (such as the 1974 oil embargo), expectations for inflation typically evolve over time and reflect the influence of monetary policy, global demand and cost-push pressures. The Fed s activity reflects the economic and monetary events shaping these expectations as seen through the prism of their statutory objectives of maximum employment, stable prices, and moderate long-term interest rates. It is the impact of these policies that is of greatest concern to investors today. It is difficult to describe the current interest rate environment as being driven by inflation expectations. The trailing five years of data on the break-even 10-year TIPS rate (showing expected inflation versus nominal yields), as portrayed in the Bloomberg chart that follows, suggest that that the market is not anticipating any significant change in the inflation outlook. To the contrary, inflation expectations have remained fairly tame and more or less range bound between 2.0% to 2.5%. These readings are consistent with current economic conditions which have been marked by sluggish job and wage growth and modest to weak overall demand. 4

5 Source: Bloomberg In contrast, the current shape of the yield curve seems to tell a different story and reflects the impact of aggressive Fed policy designed to manipulate not only short-term rates but the shape of the curve itself. Whereas, under normal circumstances, very short-term yields might be expected to reflect current inflation (1.5% as of January 2014 according to the Bureau of Labor Statistics) plus a small real return, current short rates are nowhere near this level. The artificially low short-rate and its relationship to the intermediate and longer portion of the curve is at the core of investors collective concern regarding interest rate increases following a withdrawal or reduction of the Fed s unprecedented market manipulations. Given this context it seems logical to focus historical performance comparisons on those periods not just of interest rate increases but also of periods marked by an increase in Fed activity. Accordingly we present the following 33+-year history of Federal Reserve changes (in circles) to the targeted Fed Funds rate: Source: Bloomberg 5

6 Prior to 1990 the Federal Reserve did not necessarily make fixed-increment changes, nor did they dictate these changes to the markets via direct communications. For these reasons, the first third of this chart reflects observed changes to the targeted Fed Funds rate as described in the historical records of the Federal Reserve Bank of New York. Changes to the rate made after 1990 (including intra-month changes) are reflected by month, and occur in the more readily recognized stair-step (0.25% or 0.50% increments) fashion. Utilizing the same format as the earlier data on overall changes in interest rates, the periods of greatest Fed activity with regard to increasing rates are as follows: Fed Funds: Historical Periods of Rising Rates Onset Yield Peak Yield Duration (Months) Absolute Move Yield Change % Mar % Jun % % 25.00% Dec % Sep % % 38.24% Nov %* Oct % % 26.09% Jan % Feb % % % May % May % % 36.84% May % Jun % % % Average % % Source: Bloomberg *November 1986 marked the last month preceding a gradual increase that began (according to the NY Fed data) in late December and reached a peak of 7.25% just prior to the October 1987 Crash, at which point the Fed lowered rates aggressively at first, then gradually into mid-february, Returning with these periods to the previous format for performance comparisons produces the following chart: Relative Performance During Periods of Rising Fed Fund Rates Onset Onset Yield Peak Peak Yield Reaves ERISA Composite S&P 500 Perf Diff Mar % Jun % 8.67% 1.61% 7.06% Dec % Sep % 62.19% 30.31% 31.88% Nov % Oct % 0.87% 6.40% -5.53% Jan % Feb % 13.20% 10.73% 2.47% May % May % 19.29% 18.95% 0.34% May % Jun % 69.46% 35.01% 34.44% Average 28.95% 17.17% 11.78% Yield information source: Bloomberg Notes on Chart Data: Reaves ERISA Composite (Net of fees); The S&P 500 Index is a cap-weighted representation of U.S large-cap stocks. 6

7 Both the Reaves Composite and the broad market managed positive returns over every period. Our Composite produced excess return in all but one scenario (November of 1986 through October of 1987), and outperformed the S&P 500 overall by nearly 11.8% on average. While it seems likely that the Fed will at a minimum continue to withdraw liquidity (and may eventually tighten), history suggests that when they have tightened in the past they have managed to do so without necessarily causing broad market declines. Other Potential Outcomes More Significant Rate Changes Some investors are concerned that given the unique nature of the current rate environment the expected rise in rates could be much larger, and perhaps historically unprecedented. Amidst these heightened concerns it is informative to consider what, we think, the market as a whole is expecting by way of the term structure of the current yield curve. As described earlier, the yield curve typically assigns short rates a yield approximating inflation, most recently (January, 2014) reported at 1.5%. Since 1965, in the roughly 84% of the time that the yield curve has been positive, the average upward slope of the curve (between one and twenty-year Treasuries) was +133 basis points 4 (100 basis points = 1%). We have observed that TIPS pricing suggests a forward expected rate of inflation of 2.0% to 2.5%. Using this back-of-the-envelope estimate, history suggests that an average yield curve slope over normal short-term rates might imply future 10 to 30- year Treasury yields of 3% to 4%. Most recent 10-year yields are approaching 3%, while the 30-year is just under 4%. While short rates could certainly increase significantly, from this cursory perspective, any belief that intermediate to long Treasury yields are destined to rise materially higher, however derived, is simply not shared by the broader market. This is not to say that this outcome could not happen, but as a discounting mechanism that reflects the average expected outcome of literally the world s savviest investors, it is not the market s current consensus. Short-Run Flows Clearly there has been no point in history that the Federal Reserve has been as active or has used so many non-conventional tools in an attempt to impact the level of interest rates and the shape of the yield curve. Not surprisingly these efforts have produced several (perhaps unintended) side effects. One side effect has been the decision by many income-oriented investors to abandon or reduce their holdings of fixed income and instead invest in other areas of the market. One very popular area of investment has been high dividend (and/or low volatility) stocks, an area of the market encompassing many of the names typically considered for our portfolio. According to Bloomberg, the top five U.S.- oriented high dividend ETFs currently control over $32 billion, much of it having been accumulated in the previous two years. This is not the first time this has happened; and high dividend stocks are not the only potential fixed income substitute. The same phenomenon occurring now has also occurred in the past in many areas including core real estate, Master Limited Partnership vehicles (MLPs), Real Estate Investment Trusts (REITs) and the like. Often when these short-term investors move en masse it has a negative impact on share prices. At present it is not clear where this money would move to, given the fact that cash and Treasury rates, even if increased substantially, remain very low. In anything but the shortest time frame, 7

8 we remain convinced that our emphasis on companies with earnings from typically non-discretionary products and services which have an ability to raise future dividends will help insulate our portfolios from this temporary supply/demand impact. This belief seems well supported by our performance in previous periods of rising rates. Fundamentals We often hear fundamental arguments as to why equities should perform poorly when interest rates are rising. The logic varies, but is usually associated with what has become known as the Fed Model, in reference to an obscure 1997 comment in a Humphrey-Hawkins testimony report 5. In a nutshell the Fed Model suggests that historically when interest rates rise PE multiples should fall (and vice-versa). Depending on who is making the argument, the reasons could be 1) competitive attraction (the yield on bonds versus stocks), 2) fundamental valuation (higher discount rates reduce current values) or 3) historical observation (past data supports this expected relationship). While the rising interest rates/falling PE ratio heuristic has become so widely held as to be nearly gospel among media pundits, the facts are not nearly as clear. Academics 6 routinely point out that there is no theoretical justification for the Fed Model. With regard to competitive attraction, over anything but the shortest term the comparison of the earnings yield (the inverse of the PE) to a bond yield is apples and oranges to the extent that an earnings yield is real whereas a bond yield is nominal. From a fundamental valuation standpoint the comparison effectively assumes that dividends are the only source of future returns on equity and that investors do not demand (or receive) a risk premium for investing in stocks. And finally, while certain periods in history (between 1980 and 2005 there is an observable correlation between the ten-year yield and inverse PE ratios) support the Fed Model s suggested connection in valuation, many more periods (including the past decade or so) thoroughly reject it. Summary In summary, we share concerns about the Federal Reserve s recent activity and its ultimate impact on the markets. But since our crystal ball is no clearer on the topic than anyone else s we take comfort in the fact that we are investing in an area that has historically performed well in periods of rising rates over the past 30+ years. These results may come as a surprise to the casual investor, but in reality we think they are the outcome of a carefully constructed, actively managed portfolio designed to take advantage of a rising dividend stream and include exposure to the energy sector, a natural hedge against inflation. The fact that we invest in companies with real assets that form a great share of the operating infrastructure for the U.S. economy is also comforting. Plus it never hurts to derive a good share of your returns from income since over time current income dampens volatility as evidenced by our three and a half decades of strong risk-adjusted performance. In addition, the fact that our portfolio companies often have earnings that are directly tied to inflation seems to us to be a good strategy when there is so little downside and, unfortunately, plenty of upside for future price increases. And unlike many products and services of the 21 st century, electricity, energy, telecommunications and the like enjoy significant barriers to entry, strong and unwavering demand and fewer competitive threats in what is otherwise perhaps the most volatile, dynamic and unpredictable period of economic activity in human history. We appreciate your consideration and would enjoy discussing your thoughts on this topic or in providing more detailed analysis of our investment process, our portfolios and our performance. 8

9 1 All references to Reaves portfolios, holdings and performance data are to the Reaves ERISA Composite and, unless otherwise noted, all data is net of fees. The Reaves ERISA Composite reflects the dollar-weighted return of all corporate ERISA pension accounts with assets of at least $1,000,000 under management for all periods presented (the minimum was $900,000 during the period 08/31/10-06/22/12.) Returns are time-weighted and include the reinvestment of all dividends and other earnings, net of commissions. The ERISA Composite does not reflect all of the Reaves assets under management. 2 The S&P 500 Index is a capitalization-weighted index of 500 stocks designed to measure performance of the broad domestic economy through changes in the aggregate market value of 500 stocks representing all major industries. The typical Reaves portfolio includes a significant percentage of assets that are also found in the S&P 500. However, Reaves portfolios are far less diversified, resulting in higher sector concentrations than found in the broad-based S&P 500 Index. 3 Treasury Inflation-Protected Securities (or TIPS) are the inflation-indexed bonds issued by the U.S. Treasury. The principal is adjusted to the Consumer Price Index (CPI), the commonly used measure of inflation. When the CPI rises, the principal adjusts upward. If the index falls, the principal adjusts downwards. The coupon rate is constant, but generates a different amount of interest when multiplied by the inflation-adjusted principal, thus protecting the holder against the official inflation rate (as asserted by the CPI). 4 As presented by Crestmont Research, copyright , accessed December 13, Monetary Policy Reports are mandated by the Humphrey-Hawkins Full Employment Act of 1978, which required the Federal Reserve to formally report on its activities to Congress. The Monetary Policy Reports were previously referred to as Humphrey- Hawkins reports. 6 One of the many academic papers on the topic is by Cliff Asness (who is both an academic and a successful money manager) and is entitled Fight the Fed Model: The Relationship Between Stock Market Yields, Bond Market Yields and Future Returns. This commentary has been prepared solely for informational purposes and is not to be construed as providing investment services or recommendations. Opinions and estimates are as of a certain date and subject to change without notice. Past performance does not guarantee future results. Any investments may not be suitable for everyone. An investor should consider investment objectives, risks, charges and expenses carefully before investing. For further information, please contact the Reaves sales desk at or visit our website at 9