In doing training sessions on LDI, trustees can often (quite rightly) get concerned with the complexity and the mechanics. I think this is often because LDI is looked at as a standalone product. Which it is not. As we have talked about before it is important to start from basics and understand first what you are trying to achieve before getting into the detail.
In an attempt to illustrate this, let’s use an example which will be familiar to many: buying a house.
Imagine your objective is to buy a very specific house within the next 20 years. Today that house costs £100k but you only have £50k in savings. What do you do?
The first (and easiest) solution is to hope the house falls in value by 50%, but of course most of us wouldn’t consider that a viable option.
A second, and more realistic, choice would be to model how you think the price of that house will evolve and design a savings and investment strategy that closes the gap between the money you have today and what you think the house may be worth in the future. The risks with this approach are clearly the complexity of the modelling and reliance on assumptions.
Consequently, most people in the UK would take a third approach which would be to buy the house using a combination of:
- 50% deposit
- A mortgage on the other 50%
- Salary, Savings and investment return to pay the mortgage
The first benefit of this approach is that you gain immediate exposure to the house so you don’t need to worry about the price changing. Secondly, you have more certainty about how much money you will need to save/earn in order to pay for the house in full – i.e. the mortgage payments. It is this certainty that makes this approach the preferred way of buying a house in the UK.
This is an almost identical situation to UK pension schemes.
Closed UK pension schemes have a specific objective – to meet a (relatively known) set of cashflows. Ideally, they would buy an asset that guaranteed those cashflows (a buyout). However, the average scheme only has assets totalling approximately 50% of the buyout cost.
In this case, contributions from the sponsor represent the savings part with the trustees responsible for the investment decisions required to close the rest of the gap.
There are various options available:
Option 1: Hope that the cost of buyout falls.
Option 2: Build/Buy an Asset-Liability Model that aims to model how the contributions, assets and buyout value will progress over time and to optimise a strategy that aims to close the gap.
- Buy some of the asset (i.e. have an allocation to gilts or execute a partial buy in)
- Get exposure to the asset via LDI – i.e. remove the exposure to the buyout cost going up or down
- As with a mortgage, this exposure comes at a cost in the form of a benchmark. For example, the resulting effect of utilising LDI may be that the equivalent of the “mortgage payments” are to earn a Cash+3% (for example) return on the residual contributions and assets
- Invest these contributions and assets to target this benchmark
The maths behind my examples above is near identical (house prices are more or less as unpredictable as buyout costs, for example). However, most readers would feel comfortable with the third approach to buying a house (deposit plus mortgage) yet concerned by the “complexity” of Option 3 in the pension example.
If you looked at a mortgage as a standalone financial product then it would be pretty scary. However, using it for a given tangible purpose makes it less so. The same is true for LDI – so do not look at it as a product in isolation, but as something that can be used to provide certainty over how you are going to meet your objectives.
P.S. And for those that think LDI only exists because of regulation, please note that I haven’t mentioned funding level once.