BA Pension Scheme

[john boyle]

But that's not possible!

Eh? Oh yes it is!

It is perfectly possible to perform a review in exactly the same way as
any other statutory actuarial review of a pension.!!

The usual way is to use the *difference* between future assumed investment
return and future assumed salary increases -- and similarly the *difference*
between future assumed investment return and future assumed inflation
increases.

That's because equity returns tend to be 'real' - and hence move with the
level of salary increases / inflation.

But if you are "locked-in" to a particular fixed return on gilts, that
wouldn't be possible...

So, you seem to be saying that every statutory actuarial review of a
pension fund is impossible?

 

[Tim]

Eh? Oh yes it is!

<panto time>
Oh, no it isn't!!
</panto time>

It is perfectly possible to perform a review in exactly the same
way as any other statutory actuarial review of a pension.!!

Of course - but it is *not* possible to allow for future salary & inflation
increases "... in exactly the same way as any other statutory actuarial
review of a pension ...", when the underlying future investment returns are
*FIXED* and not *REAL*.

[Because the usual method uses the *differences* in rates which don't change
much, even when absolute values do change -- but this won't be the case when
the future investment return is *fixed*.]

So, you seem to be saying that every statutory
actuarial review of a pension fund is impossible?

Not at all. See above.

Example:

Suppose we had assumed future investment returns to be (say) 7% pa, future
salary increases to be (say) 5% pa, and future inflation to be (say) 3% pa.

Now, it doesn't matter if the actual rates come out (in the future) at 11%
(investment), 9% (salaries) and 7% (inflation) -- because the differences
are still the same, so the assets value & liabilities value will still bear
the same relationship with each other (they will both have increased from
our forecasts at roughly 4% pa).

*But* if the actual rates come out at 7% (investment), 9% (salaries) and 7%
(inflation), and the assets started off at 100% of the liabilities, then the
assets won't be sufficient any longer to cover the liabilities (the assets
will be 4% short for each year at 7/9/7).

Do you see?

 

[Tim]

The nature of asset backed investments is exactly as you describe,
but it certainly IS possible to deduce that fixed interest securities are
the correct asset class for a fund if that fund is sufficiently in

surplus.

There is no one "correct" asset class.
But the point I have been trying to make, is that to reduce risk, it is best
to "match" the assets to the liabilities - eg if liabilities are *fixed*,
then invest in *fixed* assets; or if liabilities are linked to inflation,
then invest in real assets (index-linked gilts might be best then, if you
can get enough of them!).

If you are 100% invested in Fixed Interest, and inflation / salary increases
suddenly shoot upwards increasing your liabilities, then you can soon have
problems.

With a fund closed to new members (as I think
the Boots scheme is, but I could be wrong) ...

I don't think it is; altho' AIUI, eligibility for entry is quite strict -
must join scheme before age 25 or within 2 years of joining company - so a
GPP (now Stakeholder) was set up to accept employees not eligible to join
the main scheme.

But then, even "closed" funds can have liabilities linked to salary - for
current (already joined) active members.

... then this is even easier. *Every* triennial actuarial review of
a pension fund makes certain assumptions and will predict
a minimum return needed to ensure that the fund can fulfil its
future liabilities based on a certain set of assumptions. It is
perfectly possible for this to provide a 'critical yield' that is
sufficiently low for it to be achieved from Fixed Interest securities.

That "critical yield" will still depend on future salary increases &
inflation. For example, it may be "3% over inflation". Investing in real
assets should be more likely to cover that required yield - if inflation
rises (increasing the required "critical yield") then the value of the
assets should increase too, to compensate.

But if you have "real" liabilities (linked to inflation & salary increases)
and you invest 100% in Fixed Interest (say yielding a fixed 7% pa) - then
you need to be sure that inflation won't exceed, say, 3%. In other words,
you are gambling that your opinion of the future is right....
[As I said before, Boots have been lucky...]

 

[Tim]

Well i think you are questioning the validity
of every actuarial valuation of a pension fund.

Well I'm not at all (as I said before).

What I'm actually questioning, is the validity of investing in assets which
are a **very poor match** to the liabilities.
And pointing out that this strategy is a *gamble*, and that Boots have (so
far) been *lucky*.

 

[john boyle]

Well I'm not at all (as I said before).

What I'm actually questioning, is the validity of investing in assets which
are a **very poor match** to the liabilities.
And pointing out that this strategy is a *gamble*, and that Boots have (so
far) been *lucky*.

Good. We have tidied up my original point then, i.e. that is it at least
'possible' but just not a good idea because of the risk. This is
different from being 'impossible' which was your original assertion.

 

[Tim]

Good. We have tidied up my original point then, i.e. that is it
at least 'possible' but just not a good idea because of the risk.

Talking about the decision to invest 100% Fixed Interest, yes.

This is different from being 'impossible' which was your original

assertion.

Ah, but what I said was "impossible", was your assertion that they could
allow for unknown future salary increases and inflation "... in exactly the
same way as any other statutory actuarial review of a pension ". I still
hold by this.

The common method (when investing in real assets) would be to do a
deterministic valuation using assumptions for these economic factors where
the *differences* are the important factors, not the absolute values.
[It doesn't matter if each assumption was 4% out, because the end result
will be roughly the same.]

However, when you have *fixed* assets and *real* liabilities (like with
Boots), then this method would not be appropriate (and any actuary using it
should be shot!) -- you'd seriously need to do it stochastically, which is a
whole different ball-game...

 

[Tim]

... "Tim" wrote
The common method (when investing in real assets) would
be to do a *deterministic* valuation using assumptions for these
economic factors where the *differences* are the important factors,
not the absolute values. [It doesn't matter if each assumption
was 4% out, because the end result will be roughly the same.]

However, when you have *fixed* assets and *real* liabilities
(like with Boots), then this method would not be appropriate
(and any actuary using it should be shot!) -- you'd seriously need
to do it *stochastically*, which is a whole different ball-game...

John, are you still sucking those eggs?
[No reply yet?]

 

[Tim]

[No reply yet?]

"john boyle" wrote
I didnt think your last post begged a reply!

Does that mean you agree with me now? ;-)

 

[Tim]

NO ...

Oh dear...

... you seem to be just saying the same thing over and over again.

That's because I haven't changed my mind!

As I have already said, are you 'rubbishing'
every Actuarial Pension Fund Review ?

As I've already replied several times, "NO" -- I am most definitely *not*!
But I *am* saying:-
(A) The usual method used, when assets held are matched
to liabilities, is a **deterministic** approach;
(B) It's IMPOSSIBLE to (properly) use a **deterministic**
approach when assets & liabilities don't match --
such as with *real* liabilities and *fixed* assets;
(C) I believe you need a **stochastic** approach in
that case (when assets & liabilities don't match).

However, *you* seemed to think that (B) is POSSIBLE -- how so?

With regard to Boots, I think their fund was closed to new
members and its main liabilities related to pensions in payment
and, in line with most occupation defined benefit schemes, I
would expect there to be an upper limit on annual increases.

I am fairly sure there are still salary-related liabilities (even if it
is only from current actives; but I also believe there are still a flow
of a few new entrants - I don't think it's fully closed, as I said before).

 

[Rick]

I see that this scheme has a deficit of £1.5 billion.

Blame the old employees for living too long. Selfish bastards.

Well, I guess we were smart enough to produce younger bastards. Or maybe they just did that themselves... I was smart enough to be sure I didn't depend on any 'pension scheme', and my retirement income is probably less than 20% from my pension.