E-piphany

The market seemed not to mind, initially, the weak print on Existing Home Sales. Contrary to expectations of an increase, Sales actually decreased despite the fact that the home buyer credit hadn’t lapsed yet during the period covered by the data. The National Association of Realtors suggested that bottlenecks associated with a surge in sales are responsible for the downside surprise, but several observers expressed skepticism…and I am one of them. Last month sales were a 5.79mm pace and this month, because of “bottlenecks” the pace of sales actually diminishes? (To be fair, I am not sure which way the seasonal adjustments run this month, so theoretically the actual sales in May might have been higher than in April and merely adjusted lower – but it wouldn’t be a huge effect like it would be if we were comparing to, say, December).

But as I said, the market surprisingly ignored this information and tread water for a few hours. Finally, gravity simply took over. Buyers have been unable to capitalize on the “double bottom” breakout on the charts from a week ago, and what doesn’t go up eventually goes down. At 2:20ET or so, the S&P passed back below the demarcation of the top of the previous range, and from that point the end was pretty clear. Stocks lost 1.6% on the day. Treasury futures gained 19.5/32nds (TYU0) with the 10y yield down again to 3.16%.

Although volumes continue to be tepid, the technical outlook for equities just went from neutral to decidedly poor. Bonds are looking quite bullish, and targets below 3% for 10-year yields are plausible. I think a rally back to the highs of late 2008 is quite unlikely, but there is a chance that the March 2009 highs will be challenged (10y yields around 2.50%). This is pretty challenging during the summer months, though, and I doubt it will be an aggressive rally. It may well be difficult to trade.

Wednesday’s main event is New Home Sales (Consensus: 410k, reversing a spike to 504k last time), which perhaps explains why the reaction to Existing Home Sales  today was muted: there’s a much better chance of beating the NHS estimates, blunting some of the concern about EHS. The FOMC also releases a statement tomorrow afternoon; I don’t think this will be particularly illuminating, but it will be interesting to see if some of the upbeat language from recent statements is guarded after the recent data setbacks.

(The rest of the comment is not market-sensitive insight, but concerns an issue that is more than a back-burner issue for many localities).

.

In a column in the June 22 Asheville Citizen-Times, former assistant North Carolina State Treasurer Tom Campbell points out that North Carolina’s promises to retirees from state government are unsustainable. The pension plan, once fully funded, needs $400mm currently plus higher state contributions going forward; the health care plan, like most such plans, is run on a pay-as-you-go basis despite the penalties imposed on such plans by GASB 43 and 45. In North Carolina’s case, this plan is $30bln underfunded, but this is by no means atypical.

I have done some work on the problem of hedging such plans of this type, and the answer turns out to be remarkably straightforward (which does not mean, however, that they are politically palatable answers). The problem is really two problems: first, how to hedge and defease the liability for existing retirees and employees; second, how to hedge and defease, or restructure, the plan for future hires.

The first part of that problem is often a very large one, but it is tractable. If you separate all of the existing retirees (also known as the ‘population’ of the plan) into one group, the actuarial demands are reasonably straightforward. [I abstract here from longevity risk; if all of a sudden everyone begins living much longer, then these solutions shift.] We can, using actuarial tables and the distribution of ages of the group, project the number of contracts that will likely be outstanding next year, the year after that, and so on. Obviously, this number declines over time as the population of the plan diminishes. The pension plan payments are simple and can be taken straight from plan documents. The health care part is more difficult, but with some straightforward assumptions about real consumption of medical care, we can come up with a reasonable value for the real cost of that care for the life of the plan and an approximate schedule for when these payments will be made. This is, after all, what actuaries do when they are evaluating such a plan. We can also go further and, if this is a sufficiently large plan so that actuarial uncertainties are small, mostly hedge this exposure against increases in general inflation. Let’s call the resulting portfolio the “immunizing portfolio.” It obviously involves a lot of inflation-linked bonds and/or derivatives, but it isn’t hard to do.

When we know the cost of this portfolio, we can determine how much the municipality needs to set aside in order to fund these claims. This is usually a very large number. The city, state, or other entity has two choices at this point. They can fund this immunizing portfolio, either immediately or over time, and put the problem to bed. Or, alternatively, they can take the existing assets in the plan and punt on high returns, hoping that they can outrun the problem. Over the last decade-plus, this approach has led to massive underfunding…and we’ve only considered the easy part of the plan.

So there is a solution. And, frankly, this part of the solution is tractable, although it requires a will to confront the problem. North Carolina Treasurer Janet Cowell is one of those who have actually had the guts to point out the size of the problem in dollars and cents, but many public servants simply choose to pass the problem to the next administration.

But remember, we’ve only addressed the easy part of this equation: the current retirees, which we all agree we have obligations to – but who are no longer earning benefits, and the pool of which is finite (that is, they will all eventually die and the obligation to that cohort will then be zero). The more difficult part is the claims of current employees and future employees. This problem, as it turns out, is very messy. The population is variable, in size as well as in composition (ages, etc), in contrast to the problem with current retirees where the pool is fixed. Also in contrast to the first problem, the obligations to the group of “actives” have a very long duration, meaning that the plan has exposure to far-distant medical care costs and changes in longevity, and therefore the plan is very exposed to even small changes in estimates. Moreover, this is an open-ended problem: as long as new hires are granted pension and health care benefits, the obligation grows almost without bound in nominal space. The chart below shows the pay-as-you-go costs for one particular state’s health care plan I analyzed several years ago.

It isn't the current retirees who are the problem.

I think the real challenge here is blindingly clear, and I think the solution is almost as clear. (1) Hedge the obligations for current retirees – immunize that problem, put a price tag on it, and defease it over the next 30 years. (2) Make changes in the plan for current employees that focus on cost-containment strategies. (3) Greatly reduce, or even eliminate, the plan as it applies to future hires. That is, let the nanny state stuff go.

By doing this, states such as North Carolina can at least put a fence around the problem. The more responsible states will then take steps to immunize their exposures and retire them over time; the less responsible ones will keep punting and hoping they can grow out of their problems. Personally, I think the first step to getting out of a hole is to stop digging, but bad finance theory offers hope to plan sponsors that they can get more return by taking more risk.

It may not come as a total surprise to readers that even though I have been talking to plan sponsors about solutions such as this for the last five years – states, cities, big private plans – and laying out these simple albeit stark choices in very blunt terms, I have yet to find a single one that is willing to bite the bullet and solve the problem.