KEN T: When you give a series of different measurements, you
risk the question, “Which is the right one?” And it’s dependent
on the stakeholder who is asking the question. In providing
more information, you have to provide more explanation so that
people don’t pull the wrong number and make a decision on that
basis. You also need to safeguard against some people using a
particular measurement to mislead others.
Still, the market value of liability is an interesting measurement. It says, “If I take no risk, this is my obligation.” From
there, I can build according to the level of risk tolerance that a
system might consider. For systems that have gone through the
2008 markets and don’t like the level of risk, it’s a benchmark
to help them move to a lower risk stance or budget their risks.
There’s one other downside, which is that there is no single
definition of the market value of liability. Market interest rates
change every day, every hour, every minute in relation to the
bond market that is used as a proxy. The moment you measure
it, it’s history. Too much reliance on any one measurement has
its own risks.
RIZZO: If a client asks us to calculate this for them, we are glad to
do it. But clients generally don’t want it calculated because they
know it will be either used to mislead people or misunderstood.
The market value of liability has relevance in a discussion
with decision makers concerning risk. It’s not relevant as the
official balance sheet liability of the pension fund because when
researchers go out and do research on comprehensive annual
financial statements of governments, they’re going to pull off
the balance sheet liabilities. And under the new GASB standard,
the measure of the liability that recognizes the long-term rate of
return is the one that’s going to go on the balance sheet.
PhoToS.CoM
When you’re asked to calculate something, it’s really important to understand what the purpose of it is. In a workshop
session with a client concerning risk, market value may be the
starting point of that discussion. But for the purpose of funding
the plan, it’s not relevant; for the purpose of official financial
reporting of the cost to taxpayers, it’s not relevant.
In a workshop session with a client concerning risk, market value may be the starting point of that discussion. But for the purpose of funding the plan, it’s not relevant; for the purpose of reporting the cost to taxpayers, it’s not relevant.
NORTH: I’d like to pick up on the pricing of the value of benefits—purchasing a past service, for example. One could argue
that part of the reason benefits have been improved to what
some believe are unsustainable levels is that they were priced
based on what the budgeted contribution amount would be
rather than the real value of the benefits. In a low-interest-rate
environment, the cost of a stream of benefits under a market
value liabilities approach would be greater than the funding
requirement.
Almost all fiscal analyses required by legislatures, including
mine in New York, ask for the effect on the budget. If you use
an expected rate of return, you get a certain expected change
in the budget amount. But that’s based on an expected return
on assets that implicitly puts the risk on future generations that
you will actually get above the risk-free return.
If benefits were priced on a risk-free basis, you could invest
however you wish and at least the initial level of benefits provided would be something that would be less dependent on
always hitting that investment return expectation.
RIZZO: I think that actuaries can bring a lot to the table in this
area. If a benefit proposal or an asset allocation is being considered, we should be stress testing. What is the volatility in future
contributions that you can expect by adopting this benefit or by
not adopting it? When I look at risk management, I feel the metric has more to do with the funded ratio and the contribution
volatility that the employer can sustain if they have a 70/30 mix
versus having a portfolio of all bonds. Those are better metrics.
KEN T: On the other hand, when you are talking about the cost of
benefits in lieu of compensation increases and if you are trading
current compensation for deferred compensation, then there
may be value in looking at the market value for comparison.
If you’re trading current compensation for deferred compensation that is discounted at a presumed investment return,
you could be overdiscounting the value of that deferred compensation. For public systems looking at mitigating some of
the risks they’ve taken in the past, it’s important to consider
