Funding plans and contribution schedules are often set on a technical provisions (TP) basis. This is a value broadly expected to represent the amount of assets needed today to be able to pay off all the liabilities. These bases can vary enormously form scheme to scheme, and there is a lot of flexibility in choosing the valuation assumptions; however there are certain reasonably common assumptions which create an underlying drift upwards in the liabilities and scheme deficit.
One simple example is the inflation risk premium. The argument here is that investors will only take inflation risk if there is a suitable reward, so the market-implied inflation curves will be too high. It may or may not be true; however it means that, if inflation follows the forward rates, the liabilities will be larger than anticipated. Moreover, whether inflation rises or not, one thing that is certain is that if an inflation premium is used to value liabilities, but is not used to value index-linked gilts held in the portfolio to hedge the liabilities, then the relative funding position of the scheme will worsen over time.
On a slightly subtler note, another commonly used idea is the dual discount rate. A dual discount rate uses a higher rate before retirement than after retirement. The rationale for this is that if the scheme has a longer time to maturity, then the fund can hold a more aggressive portfolio, and expect higher returns. A typical dual discount rate might have a pre-retirement rate 150bps higher than the post-retirement rate. Depending on time to retirement, using the post-retirement rate only would equate to a PV of active or deferred members around 15-25% larger, or an annual growth of around 1.3%. This matters because, for a closed scheme, as the active and deferred members creep towards retirement the PV grows at a steady and non-negligible rate.
Now there are lots of moving parts in a pension scheme, and valuation changes from effects like these can be masked by changes in interest rates, inflation or mortality assumptions. However there are problems with using a valuation that has an upward trend built into it. It comes down to what you’re trying to do. Technical Provisions (TP) are in a large part there to judge whether there is enough money in the scheme or whether (and in what quantity) contributions are needed. For that end the assumptions are typically reasonable- for example, using a dual discount rate makes sense, as you are liable to need less money in the scheme if you can reasonably expect a higher return on it.
However, when setting investment and hedging strategies there can be problems with relying on a typical technical provisions basis. At its simplest, if the TP basis is set using two flat rates, rather than a curve, then any hedge becomes vulnerable to very significant curve shape risk. Moreover, using a non-economic basis makes your goals far less clear. If your scheme is fully funded on a gilts flat basis, you can almost just buy gilts, linkers and a longevity swap, and no longer worry about investment decisions. Being fully funded on a typical TP basis is less valuable, as you still need to generate excess returns.
Crucially, on a typical TP basis the liability value has an underlying trend upwards. This means a scheme that is fully funded on such a basis will still need to generate significant excess returns. This in turn means that you can fully hedge to TP, only to find that your TP has strengthened, and the hedge was only partially effective.
Building on from this, the dual discount rate assumption also makes the liability PV in some sense implicitly dependent on the asset allocation; if the scheme invested entirely in gilts, say, then using a more aggressive pre-retirement rate would no longer make sense. But a set of liability payments are what they are- they don’t shrink simply because you hold more equities. For devising a strategy, it may be better to use an economic basis, fix your liabilities, and give credit for your asset returns with an expected return framework that accounts for your holdings.
In summary, while TP assumptions differ, they often have an upward trend built in, and if nothing changes other than the passage of time, liability PVs on such bases will rise. This can cause problems when determining investment and hedging strategies, and it may be clearer and more effective to target a stricter economic basis. Credit can still be taken for expected excess returns, but these are independent of the liability payments, and it makes sense to treat them separately when setting strategies.