Prediction? Pain
by James Moore, PIMCO
- Recent Federal Reserve activity has pushed down the long end of the yield curve, spiking the present value of plan liabilities and widening the funding chasm.
- The pain of the pension community shows up most obviously in funded status estimates.
- High and increasing levels of implied equity risk premium in pension plans suggest sponsors’ expectations are increasingly optimistic about future contributions from risk assets.
The Rocky series is a modern update of the familiar Horatio Alger tale of the underdog triumphing over adversity. Part of what made Rocky Oscar-worthy was that Rocky ultimately does not win. Of course this sets the stage for Rocky II, where the hero triumphs over his opponent, adversity and his own foibles to win the championship. Rocky III brought us a champ who had become complacent, overindulgent, and a bit lazy, and who is savagely beaten down by a new, hungrier challenger, Clubber Lang.
There was perhaps no better villain in the eighties than Mr. T’s chiseled, mohawked, snarling Clubber Lang, and few better in movie history. Clubber Lang launched Mr. T as a cultural icon and there are relatively few so distinctly of that era who have endured so long and are still so recognizable. In October 2010, Mr. T made a memorable appearance on Bloomberg Television, extolling the virtues of his favorite investment – gold. Nearing sixty, he is no longer the fearsome presence he once was. Thirty years of wearing a golden yoke of up to 45 pounds and five years beating cancer will do that, but he looked quite good for his age and those following his investment advice would more than likely have benefited. Gold is up more than 15% since his appearance (despite the recent pullback), according to COMEX, far outpacing the average pension asset return (represented by the Milliman 100 Pension Funding Index), which has been slightly negative over the same period.
Unfortunately, pension plans are currently experiencing something synonymous with Mr. T’s Clubber Lang character. Pain. The bearded villain this time is not Mr. T, but Dr. B. – Ben Bernanke. The pain comes only partially from the asset side; in addition, the Fed’s new Operation Twist (selling short-term Treasuries in exchange for longer-term) has pushed down the long end of the yield curve, spiking the present value of plan liabilities and widening the funding chasm. By our estimates (based on Credit Suisse pension data for the S&P 500 universe), the drop in Treasury yields combined with equity market declines added roughly $80 billion in underfunding in the few days following the announcement of the policy. This brings the cumulative underfunding of corporate pensions to something north of $400 billion.
The pain of the pension community shows up most obviously in funded status estimates from equity strategists and consultants, which give mid-year forecasts based on companies’ 2010 10-K data and market changes over the year to date. As past is prologue, the funding hole gives us our starting point and helps set our course for the direction forward. Deficits imply contributions and possibly changes of course that reflect new circumstances. One would hope greater awareness and visibility would lead to better investment strategies and tighter asset-liability management. While this has been the case among a growing number of plans, overall movement has been slow and the asset allocation strategies have in aggregate not changed all that much. Data for S&P 500 companies going back to 1994, when the SEC clarified the FASB (Financial Accounting Standards Board) discount rate guidelines, reveals some interesting patterns. Figure 1 shows the Moody’s long-term AA index – which is generally a good proxy for the average plan discount rate – along with the year-end 10-year Treasury yield and the difference between the average assumed plan return (source: UBS and Credit Suisse data on long-term expected plan returns) and the discount rate proxy. The pattern for the two yield series is abundantly clear and well known: Bond yields have fallen over the past 17 years. The arc of the long AA index series is smoother than the Treasury yield series largely because credit spreads are typically countercyclical. Treasuries spike down on flights to quality while spreads tend to rise.
The plan asset return vs. discount rate spread tends in the opposite direction, rising from 0.78% in 1994 to 2.64% at the end of 2010 and we expect it may trend higher through 2011 given recent market activity. This is because the assumed returns have fallen more slowly than discount rates. For 1994, plans’ expected long-run return was 9.4% (source: Credit Suisse). Over the years this has trended down to an average 8%, and we believe it will drop further by the end of 2011. The rising spread indicates that plans expect their risk assets to work harder and carry more now than they have expected historically. It seems paradoxical that expectations should be greater for contributions from risk assets going forward when they have disappointed so much over the past decade.
Given the data above and the pension plan asset allocations presented in sponsors’ annual reports, we can derive a sponsor-assumed equity risk premium (ERP) over the 10-year Treasury. Figure 2 shows the implied equity risk premium values estimated for the S&P 500 plans in aggregate (based on Credit Suisse data). The time series mirrors the asset-discount spread rising over the past 17 years. What is more notable is the degree to which it rises and the ERP currently implied. The implied ERP rises from a low of 2.4% in 1994 to a previous high of 9.6% in 2008. Assuming rates hold their current levels and aggregate S&P 500 plans’ expected returns drop to 7.5%, the current implied ERP would be 11% (as of September 2011). If expected returns instead remain unchanged from year-end 2010, the implied ERP would be 12%.
Aside from the extremity of these implied ERP levels, note the pattern of volatility. The equity risk premia display local peaks in 2002 and 2008 at the depths of recessions and local troughs in 1999 and 2006 at the end of cyclical expansions. This is consistent with both the countercyclical nature of the 10-year Treasury yields and the notion that equities (represented by S&P 500) tend to be cheaper at times of great strife and despair and tend to be richest after runs of prosperity.
More worrisome is the rise in the implied risk premium and its current level. For comparison, we can use Pastor and Stambaugh’s estimates of equity risk premia from 1834–2000: They find a range of 3.9%–6.0% over that long period. Current estimates for consultants and asset allocation practitioners vary quite a bit depending on models and biases, but are largely spanned by a 3%–7% range.
What do these levels imply? In the early nineties, plan sponsors, if biased in their forecast, were generally biased toward conservatism. From 1997 through 2007, expectations, although a bit rosy at times, were largely within the realm of reasonableness. In our view, a long-run equity risk premium of 11% is pure jibber-jabber. It is wishful thinking. I dare not predict the level of the S&P 500 ten years out, but an ERP this high suggests the S&P would have to reach unprecedented levels. If this is what plan sponsors are counting on, I, like Clubber Lang, predict Pain.
Rocky III was released Memorial Day weekend, May 1982. Times were tough – real tough. The U.S. was in the depths of recession. Inflation was at 7% and the 10-year Treasury was priced to yield 13.7%. Unemployment was at 9.4% and heading higher. Part of the reason Rocky III resonated with its audience and was a box office smash was they too were beaten down and overcoming obstacles to rise against adversity.
Rocky rebounded from his loss to Clubber Lang, but he did not rely on hope in preparing for his rematch. Hope is neither a training plan nor an investment strategy. Anyone familiar with the Rocky series’ ubiquitous training montages will know the keys to success are hard work, sweat, an adrenaline-firing score, and some measure of pain. The answers to our pension problems are not all that different. I pity the fool who expects an easier answer.
Copyright © PIMCO
