Accounting for Pensions - The Basics

Pensions are a famously dry topic, known to send even the most seasoned of accountants to sleep. Add in their complexity and you’ve got a recipe for a poorly understood aspect of financial statement analysis. Regardless, they can have a profound effect on a company’s balance sheet (and therefore its valuation), and so it is well worth taking the time to understand them. This article is the first of a series and is intended to outline the utmost basics of pension accounting. In later articles, I’ll be looking to delve a bit deeper into pension plan analysis and then apply the knowledge with a case study.

A number of users on the site have requested that we try to incorporate analysis of pensions into our StockReports and we appreciate that many consider it a very important topic. So, with any luck, we’ll get a bit of a conversation going about how the more experienced investors among us think about pensions too. Long term, this is going to shape how we incorporate this kind of information into the site so please do get involved!

The Basics…

In the most basic sense, pensions are essentially a savings agreement between a company and an employee. Employees contribute a small part of their salaries towards a communal pot (a pension fund) and pay annual fees in exchange for the money being managed and invested for the duration of their working lives. Often, the employer also agrees to contribute a certain amount to the employees pot when they pay in too.

When the employee finally retires, the accumulated pension fund is then used to pay them an ongoing retirement benefit. Pensions are the way almost all of us fund at least part of our retirement and they are highly prevalent in the UK - as I’m sure the vast majority are you are aware of from your jobs.

Different Types of Pension Fund…

Broadly speaking, there are two types of pensions you will come across. The first is called a defined contribution pension (these are the easy ones!). The second is called a defined benefit plan (these are not so simple I’m afraid).

A defined contribution plan is where an employee makes a predetermined payment (contribution) into the pension fund every month. It has this name because the ultimate value of the fund is determined only by what has been contributed to the fund over its lifetime. There is nothing in the deal which says how much you are going to get out of it at the end.

Because the employer has not obligation to pay the employee any prespecified amount, we say that the investment risk is borne by the employee. It is the employee who bears the risk and rewards of investing in the pension fund and they only get out what is left in it at the end. Even if the pension fund manager performed extremely badly, and there was nothing left in the fund, the employer would owe you nothing beyond what was in the fund.

As such, for us investors, defined contribution plans are more or a less a non event. The only significant way they affect the company we are reviewing is how much the company contributes to the plan. This is still worth taking note of - a big company with a big defined contribution plan is going to pay out a lot in contributions, but this will be a relatively small and consistent figure. It should always come in as a fairly small % of total salary costs.

The second, and more interesting type of pension is a defined benefit pension scheme. In these plans, while the employee does make contributions throughout their life, the employer promises up front how much the employee will be paid in retirement. This is often defined by a simple formula along the lines of final salary * x% * number of years of service.

Because the employer has agreed up front how much they have to pay out, they have a certain (and potentially large) liability due at some point in the future. It is up to them to invest pension plan assets (built up through employee and employer contributions) wisely enough to meet the liability they will have in the future. As such, we say the investment risk lies with the employer...and this is where it gets interesting for shareholders.

Surpluses and Deficits...

You may have gathered that there are two distinct balance sheet components to a company’s defined benefit pension scheme. The scheme’s assets (known as plan assets), and the scheme’s future liabilities (plan liabilities).

Whenever a pension plan has more assets than it has liabilities, we say it is in surplus, and whenever it has more liabilities than it has assets, we say it is in deficit. Accountants being the confused people they are, you will often hear some similar terms being referred to - funded and unfunded status.

A funded (in surplus) pension plan is one that has enough assets to cover its liabilities - it will be able to meet all of its obligations from the plan assets. An unfunded (in deficit) plan does not however - it will be able to meet its obligations from the plan assets there are for the time being, but at some point it will have to fund the difference from its current earnings.

This is the first point to note for investors looking at companies with significant defined benefit pension schemes. Is the plan unfunded? If it is, well, the extra assets have got to come from somewhere. Either the plan manager has to earn fantastic market beating returns (unlikely) or the company will one year have to cough up shareholder’s hard earned cash to meet the liabilities (very likely).

Hang on though, all of these liabilities we are talking about, aren’t they future liabilities? Current plan assets seem easy to value - it is just the market value of all the assets (stocks, bonds etc) in the portfolio…but future liabilities? How do we know what they are worth? Good question, and, as you will see, it is the question when it comes to pension analysis.

In step the actuaries...

It is the job of actuaries to figure out the value of these future liabilities, so how do they do it? There are a few key components to the process. The first is to estimate how long, on average, your employees will work for you. As we saw in the formula above, retirement benefits are often defined by how long an employee has worked for the company so this is one of the first key steps to working out how much you will be paying out every year.

If the average age of plan entrants in your defined benefit scheme is low, and staff turnover is also low, then you can predict that you will have a high average number of years of service by the time the plan matures. On the other hand, if, on average, people join the plan late in life and don’t stay for very long at the company, then you can predict that the average number of years of service will be low. Once you’ve done this you now need to forecast how long they will live for.

Actuaries use mortality tables to estimate the probability that a given person will live to a given age. When they combine this information with the average retirement age of a workforce, it is fairly easy to calculate an average number of years of retirement benefit that will be due to current employees. The final step is to discount the sum of all of these retirement benefits back to today to give their present value using an appropriate discount rate. The discount rate is chosen to reflect the rate of return the pension plan expects to earn on its portfolio to account for the time value of money. It is this final present value figure (well, not quite, as we’ll see later) which goes on the balance sheet as the liability. Easy, what could possibly go wrong with all of that?

As mad as it seems?

To those who aren’t familiar with the process, it probably seems absolutely ridiculous that anyone could hope to obtain anything remotely resembling an accurate answer from such a forward looking, assumption heavy process. Do have some faith though - it’s amazing how well these things can be estimated when the calculations are performed on a large enough scale thanks to the law of large numbers.

Take the example of Robert Wallace and Alexander Webster, arguably the world’s first actuaries. They set up the first pension fund for the benefit of the widows of Scottish ministers (yes, this was the start of what is now the pensions giant Scottish Widows) in the early 18th century and made a remarkable prediction. They estimated that 20 years from its foundation, the value of the funds assets would sit at £58,348. This was a sum that required the estimation of the future number of ministers, the number of children they would have on average, how many would die and leave widows in that 20 years and many more factors. In fact, 20 years later, the funds proved to sit at £58,347 - just £1 out from their original estimate! If it was possible for these two men to perform such an incredible feat of calculation with little more than pen and paper, then perhaps some of the best guesses of actuaries today aren’t so bad.

And then the accountants...

The problems arise, generally, when the accountants start to get involved. In their infinite wisdom, accountants across the world have decided that there are certain standards that should be applied to these calculations. We’ve just seen with the Wallace and Webster example that (to a degree) actuaries can actually make relatively accurate forecasts about the value of pension liabilities going forwards, so why do accountants have to step in and change it all up?

One reason is to help with comparability between companies which is one of the 4 enhancing qualitative characteristics of financial information as defined by the International Accounting Standards Board (IASB) in their conceptual framework for International Financial Reporting Standards (IFRS). Another reason is to make it harder for company actuaries to manipulate their balance sheets by using inappropriate or overly optimistic assumptions about discount rates, mortality rates etc. These seem like good reasons and, in some regards, they are. It was only very recently, for example, that managers were allowed to bake in an assumed percentage of market outperformance on their portfolio in order to calculate a discount rate. It’s pretty obvious what the temptation is there...

Maybe the rules are for the best then, if we can’t trust reporting managers to be honest, but they do create an interesting complication for us investors. Where liabilities have been discounted using restrictive criteria, there is a good chance that their real present value is actually quite different from the amount shown on the balance sheet. If they are smaller than presented then that’s ok, prudence is important when performing valuation work but, if they are larger than presented...then you have a problem.

Conclusion

We’ll conclude there for now before it all gets to be too much! To summarise what we’ve covered:

• There are two categories of pension - defined benefit and defined contribution.
• As investors, we are only really interested in defined benefit schemes as it is for these schemes only that the company bears the investment risk.
• There are two broad balance sheet components to defined benefit schemes, plan assets and plan liabilities.
• When assets are greater than liabilities we say the plan is in surplus, and when liabilities are greater than assets we say the plan is in deficit.
• Plan assets are the market value of the plan’s portfolio (cash, stocks, bonds etc). Plan liabilities are the best actuarial estimate of the present value of the retirement benefits due to employees in the future.
• While actuaries can make surprisingly accurate predictions about the future given a free reign (and with the assumption of honesty!), accountants impose certain standards for reporting purposes. 
• They do this in order to improve comparability of financial statements and to make it harder for management teams to intentionally misstate their liabilities.
• While this is done with good intentions, it does lead to complications for us investors because stated liabilities may not reflect reality!

That’s all for now then, I hope I’ve explained these basic points clearly and that everyone is now on the same page. For the next article, the plan is to look in more detail at the specifics of actuarial assumptions and the accounting restrictions placed on them. That should set us up nicely with the knowledge we need for a full blown case study in the third article.

Please do get involved in the debate below with any questions or ideas you have relating to the topic - particularly if you have any suggestions of what you might like to see covered in the next article. If you think I’ve explained something badly or want clarification on a point then I’m very happy to help.