As corporate treasurers get ready to publish their 2012 annual reports, we ask for one New Year’s resolution: get real on pension liabilities. Creeping pension liabilities are an increasing source of concern for credit investors and full disclosure of the risks surrounding them is what credit investors need to regain confidence in the most affected issuers. A few companies have stepped forward in recent years with some admirable improvements in their disclosures, but in general the information available to investors is still far from what we need.
Time and again, we’ve seen that credit markets penalize issuers for lack of disclosure, with spreads ultimately higher than they would be if issuers just came clean on the risks. Yes, companies can generally keep some investors in the dark for a while, and enjoy lower spreads in the short term as a result of limited disclosures. Ultimately, though, credit investors wise up to the problem and in the end are forced to assign a steep uncertainty premium to issuers.
If we have learned nothing else in the last five years, it’s that, at least over time it pays to take your lumps up front by providing full disclosure around key and emerging risks. Subprime, European peripherals, oil drillers in the Gulf of Mexico – these have indicated that full disclosures have paid off with lower credit spreads over the medium term. In 2013 and beyond, we expect pension liabilities will be a similar concern demanding similar levels of disclosure. With this in mind, we propose the following as companies develop their New Year’s disclosure resolutions:
Provide more granular information on the assets and, especially, the liabilities sides of the pension balance sheet. Despite some recent improvements, disclosure is still very scant for many companies. Many simply disclose the pension liability and maybe a very rough breakdown of plan assets, but little more other than the assumed asset return and the discount rate for pension liabilities (more on these below). Very few companies go further on the liability side and give summaries of the actuarial information they use to generate the gross liability. Public disclosures reveal little about the assumed distribution of cash flows related to these liabilities over time, the assumed mortality of plan members, the assumed end-benefit amount, the breakdown of inflation-linked and true-fixed liabilities (admittedly not material for most plans), and many other key items. On the asset side we may have a little more information from some companies but still not enough to come to any real understanding of the nature of pension plan holdings and what investors can expect about the future returns and volatility of returns of those assets. The recent move to show the break-out of assets by Level 1/2/3 groups based on certain valuation inputs is a positive but still not especially helpful as granular details on the credit quality and underlying asset classes of the assets is far more important.
Treat investors like grownups – don’t assume we can’t handle the information. A common refrain that we hear from management teams is that companies already disclose so much information that any more would simply be too much to handle, especially complicated actuarial information on pension liabilities. This is simply wrong. You cannot find a PIMCO credit analyst who thinks such additional information would be too much to handle, nor would you find that response from any serious professional credit analyst. As with many new areas of disclosure, there could be a transition during which there may be some short-term volatility in stock prices and credit spreads, but robust engagement of the investor community to educate analysts on the disclosure could keep this transition period short. If some banks and insurance companies have managed to greatly increase their disclosure of complicated information in recent years and been rewarded with better share prices and credit spreads, why shouldn’t companies enjoy a similar benefit from pension disclosures?
It’s not just about the liability – it’s about the volatility of the liability. Liability stress tests are essential, as my colleagues Sebastien Page and Rene Martel have written (see “Asset Allocation for Pension Plans: Managing Risk in the Not-So-Long Run”). Companies should know (and disclose) not just the pension liability under certain normalized assumptions, but also how volatile the pension liability could be over time. Because companies’ ability to make cash injections into pension plans is generally negatively correlated with the size of the required cash injection from one year to the next, the volatility of the liability is a critical factor in credit analysis. Such analysis at PIMCO is very much about understanding the volatility of the enterprise value of a firm – something we simply cannot do without understanding how a pension liability will behave in a stress scenario. Again, more disclosure on the duration of liabilities could help significantly here, as investors could in theory do their own stress tests with more duration disclosure.
Get real on the assumed asset return. This is perhaps one of the most difficult but most pressing changes that needs to occur in the way companies deal with their pension liabilities. As of the end of 2011 (U.S. companies typically update their pension assumptions only in their annual reports), the vast majority of S&P 500 companies with pension liabilities used what PIMCO would politely term “optimistic” assumptions of returns on plan assets. And after 2012, those end-2011 assumed rates of return will look even more aggressive. Yes, plan assets may have had decent returns in 2012 on the back of further compression in yields and the quantitative easing-led rally in risk assets, but as many PIMCO portfolio managers have written this year (see “Cult Figures” by Bill Gross) returns for stocks and bonds going forward will likely be in the low-to-mid single digits over the long term.
Most U.S. corporate pension plans, however, are still clinging to much higher assumptions of long-term returns. At end-2011, for instance, the 50 companies among the S&P 500 with the largest pension liabilities as a percentage of EBITDA had a median asset-return assumption of 8%. We recognize that under U.S. accounting rules (GAAP), companies may use historical rates of return to inform their forward return assumptions, but that still does not address that companies may be understating – and potentially materially understating – their true pension liabilities by using overly flattering asset-return assumptions (i.e., if actual asset returns are less than anticipated, then the company may eventually need to increase funding to pay liabilities). Once again, more disclosure could help here: companies should spell out in more detail how they arrive at their return assumptions by asset class.
Get real on the liability discount rate. Here again, there will need to be further declines in the rate companies use to discount their future pension liabilities. The median discount rate used by the 50 S&P 500 companies with the largest pension liabilities as a percentage of EBITDA at end-2011 was arguably already too high at 4.77%. Since then, of course, long-dated yields have only declined further, with the Moody’s AA corporate yield down another 25 bps. Yes, U.S. GAAP allows companies to use the last few years’ average yield on highly rated corporates, not simply the most recent print. But to use a backward-looking discount rate that reflects economic conditions that are likely not consistent with forward expectations is again to understate the pension liability. In line with publishing more granular data on pension liabilities, it would be extremely helpful to see what the change in pension liabilities would be if the yield on highly rated corporate bonds as of end-2012 were applied instead of the longer-term average. In addition, there should simply be more disclosure on how the discount rate is generated, especially in the cases where plans use bespoke “above median” curves from actuarial consultants where the discount rate uses a “trimmed tails” approach that often results in a higher discount rate as a result of trimming more of the lowest yielding bonds than higher yielding from their sample.
Final New Year’s resolution: Recognize that this is a real problem. Aside from providing more details and using more realistic assumptions, companies need to begin to internalize that swelling pension deficits are a serious problem, one that needs to be addressed now – by scaling back share buybacks and dividends and injecting more cash into pension plans. We understand that to date there has been insufficient market discipline around unfunded pension liabilities. Ratings agencies have been mysteriously quiet on growing pension liabilities; they appear fixated on debt liabilities rather than total liabilities – with even less attention paid to how big unfunded liabilities would be with more realistic assumptions. Equity investors have been annoyed by the chronic and volatile earnings hits from annual pension liability recognition, but here again there has been little attention paid to how enterprise value shifts from shareholders to pensioners as unfunded liabilities grow. And many credit investors have been similarly negligent, following ratings agencies’ methodologies without fully incorporating unfunded pension liabilities into credit spreads.
While market discipline may have been lacking in recent years, the problem is real and growing, and eventually the discipline will come. Much attention has been paid to the decline in net indebtedness of large cap U.S. companies in recent years, but when the rise in unfunded pension liabilities is included in the tally the decline is much less impressive. Over the four years ending in December 2011, net debt among companies now in the S&P 500 fell by $1.39 trillion to $4.13 trillion. Over the same period, however, the aggregate pension plan position of these companies swung from a $49 billion surplus to a $353 billion deficit. Over the period, the number of companies now in the S&P 500 with unfunded pension deficits jumped from 222 to 324. While it is impossible to say what applying more realistic assumed asset-return and liability-discount rates would do to this number, it would almost certainly increase unfunded pension liabilities significantly, further diminishing the improvement in companies’ balance sheets. Finally, the improvement in U.S. corporate balance sheets has already been moving in the wrong direction in 2012 even without the erosion in pension balances; net debt reversed its downward trend, rising back to $4.21 trillion as of the third quarter, and the flurry of one-time dividends in the fourth quarter will likely push that higher still.
Given how serious the rise in pension liabilities has been for many U.S. companies, the message should be clear: investors need more disclosure, and they need more aggressive and realistic strategies for addressing these rising pension liabilities sooner rather than later. By tackling the problem now, the New Year may be painful at first, but over time companies should enjoy lower credit spreads (and, ultimately, a lower equity risk premium) by giving investors confidence that the pension problem is settled.