Liability Driven Investment
5b. Monitoring such a structure: Swap
portfolio structure
[this page | pdf | references | back links]
Return to Overview
Next page
5.4 The choice
of numeraire (e.g. whether the cash flows are in nominal or real terms, or if
they are expressed using some present value metric) is not particularly
important as long as the cash flow analysis ultimately precisely specifies the
assumed cash flows. This explains which cash flows as per 5.3(d) are more problematic –
they require lots more detail to specify precisely. It may be possible to
develop suitable approximations that simplify them into a form that was more
easily specifiable. It might also in practice be possible to simplify away
liabilities of the form described in 5.3(c). It is also worth noting
that the cash flows are not deterministic in nature. If the numbers of members
involved is quite small then the random incidence of individual deaths will
introduce uncertainty. For more sizeable schemes, the unpredictable nature of
future changes in general levels of longevity is likely to be more significant
(as is whether the mortality table in question is suitable for the actual type
of individuals represented by the scheme membership).
5.5 Once the
liabilities have been expressed in a suitably simplified form it becomes
possible to structure swaps that capture the main characteristics of these cash
flows. Liabilities that are fixed in nominal terms would be matched using swaps
that generate fixed cash flows whilst those that are RPI-linked would utilise
inflation swaps. LPI-linked liabilities can be catered for in a similar fashion
although often their costs seem high to clients. This seems to be because
clients worry less than the market as a whole does about the possibility of
inflation becoming negative.
5.6 Performance
(and risk) measurement and attribution of the swaps portfolio can then also be
carried out by reference to the simplified cash flows, discounted (probably) at
swap rates, versus mark to market movements in the value of the swaps.
5.7 There is a
link between liability driven investment and fair valuation principles. The
actuary will typically have placed some value on the liability cash flows.
Assuming that the liability cash flow projections are truly correct (and
ignoring some of the niceties surrounding credit risk on cash deposits etc.),
we might ask how we can tell if this sum would actually be sufficient to
provide all of the projected cash flows. This depends on whether the actuary’s
valuation is bigger or smaller than the fair value of the liabilities derivable
from the mark to market value of the swaps. It is not sufficient merely to
compare the return on the liability driven portfolio with the movement in value
placed on these liabilities by the actuary. The movement needs to be unbundled
into its various parts, including potentially a part relating to the difference
between the fair valuation and the actuary’s valuation.
5.8 Even the
above analysis involves simplifications. For example, there is an implicit
assumption in the above that the fund’s mortality experience can be well
predicted at outset. But merely differentiating between nominal, real and LPI-linked
increases provides no protection against unexpected improvements in mortality.
There may be future discretionary benefit improvements. Active members’
liabilities are particularly difficult to project reliably in this context
given their sensitivity to uncertain future member-specific salary increases.
For a full picture one would in principle differentiate between each such risk,
as per Section 4. In practice
this is likely to be challenging, although at least thinking about such matters
may help to highlight what sorts of risks a liability driven investment
portfolio does or does not hedge against.
NAVIGATION LINKS
Contents | Prev | Next