High yield value investing

“If an insurance company consistently makes losses on its operations, the company will sooner or later be unable to fulfill its part of the insurance contracts.”

Karl Borch, The Economic Theory of Insurance

Insurers

How insurers are different

Like banks, insurers don’t use wood, steel or people as raw materials, they use money. Also like banks, insurers act as a link between those who have money and those who need it. In this case, the link is between those who are exposed to a risk (such as a car crash) and those who have been impacted by the crystallisation (i.e. occurrence) of a risk (again, such as a car crash).

In practice this means customers (the assureds or insureds) pay premiums to an insurer. These premiums then sit on the insurer’s balance sheet as assets which are sometimes known as insurance float. When an insured event occurs, the insurer takes some of that float and pays it out to the claimant. Because the insurer is liable to cover the cost of future claims, these potential claims are a liability, and the total estimated amount of future claims sits on the insurer’s balance sheet as its primary liability.

This business model means that most insurer’s don’t report revenue. Instead, their income statements start with earned premiums and net earned premiums. Net earned premiums are premiums net of reinsurance, where reinsurance is when an insurer pays premiums to another insurer (the reinsurer) to cover some portion of insurer’s risk. So as with banks, we need a replacement for revenues.

Like banks, insurers also have unusual balance sheets. That’s because insurers have huge amounts of premium assets and claim liabilities on their balance sheet compared to their net assets or shareholder equity. This makes insurers very sensitive to changes in the value of their assets and liabilities, so we’ll need some additional safeguards to take account of that.   

Updating Growth Rate for insurers

Turning to insurers, we have a similar situation. First of all, insurers don’t report revenue, so we need a revenue substitute. The top line of most insurers’ income statements shows premium income rather than revenues, so we could use that as a replacement. Another option is to look at the balance sheet rather than the income statement.

An insurer’s assets are mostly made up of premium income which is being held in reserve to cover the cost of future claims. And, perhaps unsurprisingly, an insurer’s liabilities are mostly made up of the estimated cost of those future claims. An insurer only truly earns a profit once a policy has lapsed and no claim has been made. At that point, the reserved premium can be released and used within the business or returned to shareholders as dividends.

This means the future earnings of an insurer have as much to do with its total premium reserves as they do with the amount of premium income generated each year. And if we take the view that reserved premiums are the lake from which future earnings and dividends will flow, then it makes sense to replace revenues with an insurer’s total premium reserves.

Since premium reserves are typically the largest asset on an insurer’s balance sheet by some margin, an easy way to measure premium reserve growth is to measure the growth of an insurer’s total assets. This is, of course, exactly the same adjustment we used for banks.

And the similarities continue with lease-adjusted capital employed. Like banks, insurance companies don’t need lots of property, factories, vehicles, machinery of other expensive capital assets. All they really need is some underwriters and customers who trust that they will pay out in the event of a claim.

This means that, in general, insurers don’t need to borrow lots of money to grow. In fact, rather than needing to borrow cash to grow, insurers are virtually drowning in cash as they grow. That’s because customers pay cash premiums to the insurer up front, and claims (if they occur at all) are only paid out much later on. So most insurers don’t use much in the way of operational borrowings.

This is handy because an insurer’s operational borrowings are often buried in the accounting notes, rather than clearly displayed on the balance sheet. This isn’t because insurers are trying to hide their debts; it’s just that they have various forms of borrowings, some of which is used to strengthen the capital buffer which is used to strengthen the company’s ability to pay out claims through thick and thin.

So again, as with banks, I usually only look at borrowings for insurers in the current year in order to calculate their Debt Ratio. For Growth Rate, I use lease-adjusted net assets (which doesn’t take account of operational borrowings) instead of lease-adjusted capital employed.

For insurers: Total assets are mostly made up of premium reserves. These are premium payments from customers which are held in reserve to pay for future claims. If no claim is made, the premiums can be released as earnings which can then be retained within the company or paid out to shareholders. So as with banks, an insurer’s total assets are the ultimate source from which all future earnings and dividends will flow.

Additional ratios for insurers

The additional ratios for banks focus on the balance sheet, and for good reason. We’ll also look at some leverage ratios for insurers, but first, here’s a ratio which looks at the profitability of an insurer’s core underwriting business.

The combined ratio for non-life insurers

To understand the combined ratio, we need to look at the two ways in which insurance companies make profits:

1. Underwriting profit – Writing insurance policies where, in aggregate, the collected premiums more than cover the eventual cost of claim losses and underwriting-related expenses.
2. Investment profit – Premiums are collected up front and claims, if made, are paid out later. This leaves insurers with a pool of collected premiums sitting on their balance sheet. Most insurers invest this insurance float and record any investment gains as profits (or losses if the investments do badly).

In general, a good non-life insurance company will make a profit on both the underwriting and investment side of its business. However, of the two, underwriting profit is more important because it represents a profit from the insurer’s core insurance business rather than from the ups and downs of the investment markets.

The combined ratio is how we measure underwriting profit. It’s called the combined ratio because it combines two fundamental ratios: The loss ratio (the ratio of expected claim losses to premium income) and the expense ratio (the ratio of underwriting expenses to premium income).

Loss ratio = claim expenses / premium income

Expense ratio = underwriting expenses / premium income

Combined ratio = loss ratio + expense ratio

If the combined ratio is below 100% then the company made a profit on its underwriting operations. If the ratio is above 100% then it made an underwriting loss.

If I’m going to invest in an insurer then I want its core underwriting business to be consistently profitable. More specifically, I want the underwriting return on premium income to be at least as good as the return on sales target that trading businesses need to exceed. My return on sales target is 5%, so I want to see an underwriting margin of 5%, which means a combined ratio of less than 95%:

Defensive value rule of thumb

Only invest in a non-life insurance company if its five-year average combined ratio is less than 95%.

For life insurers, the combined ratio doesn’t make much sense because their their underwriting returns are often negative. Making an underwriting loss may seem like a bad idea, but life insurers typically expect to make up for any underwriting losses with long-term investment returns.

This may seem back to front, but life insurers write very long-term policies such as life insurance and annuities, where premium are received years or even decades before claims are paid out. Those premiums can stay invested for many years, which produces larger investment returns, and if the company is run well then those returns should more than offset underwriting losses.

Having said that, some life insurers also write non-life policies, so it’s usually worth looking for the combined ratio in life insurance annual reports too, just in case.

Let’s take a look at the combined ratio of a real company; in this case, Admiral insurance.

Admiral Group’s combined ratio

Admiral is the UK’s leading car insurer and is listed in the FTSE 100. It has more than 4 million UK car insurance policyholders, almost a million UK home insurance policyholders and another million policyholders in its international car insurance operations. The company also owns several leading price comparison websites, including confused.com in the UK.

Admiral grew rapidly from its founding in 1993, helping to disrupt what was previously a broker-based car insurance market with the direct model pioneered by Direct Line and Churchill Insurance (Admiral’s first and long-time CEO Henry Engelhardt was one of Churchill’s founding team).

Unlike most of the ratios in this book, the combined ratio is an industry standard measure and as such, is quoted in the annual results of all non-life insurers.

This makes gathering the necessary data much easier, and Table 6.1 lists both Admiral’s combined ratio and its component loss and expense ratios.

Year	Loss Ratio	Expense Ratio	Combined Ratio
2014	67.80%	18.70%	86.50%
2015	65.10%	20.50%	85.60%
2016	72.00%	22.40%	94.40%
2017	66.20%	21.50%	87.70%
2018	66.40%	22.90%	89.30%
Average	67.50%	21.20%	88.70%

Table 6.1: Admiral’s underwriting ratios to 2018

Admiral specifically targets (and has long achieved) an industry leading combined ratio, so don’t expect to see many insurers with an average combined ratio of less than 90%. In most cases, my rule of thumb of less than 95% will be more than hard enough to achieve.

The premium to surplus ratio

Let’s turn to the balance sheet. For insurers, their main asset is the pool of premiums collected from insureds and reserved to cover the cost of future claims. Their main liability is the expected cost of future claims.

As insurance companies have an obligation first and foremost to fulfil their insurance contracts and pay claims promptly, their premium assets must always exceed their expected claim liabilities. The surplus of assets over liabilities is also known as the premium surplus, as it is effectively the surplus of premiums received over expected claim payments. This surplus is very similar to the capital buffer that banks use and it is required for more or less the same reasons.

Here’s a quick example:

If an insurance company has £100m of pooled premiums and £99m of expected claims it would have a premium surplus of just £1m. If the value of its pooled premiums decreased by just 2% to £98m then the company would be technically insolvent.

Why would the pool of premiums decrease in value?

The reason is that premiums are often invested to some extent in volatile assets such as equities in order to boost the company’s underwriting profit with an additional investment profit. However, as we know, equities can go down as well as up. With a 2% decline in the value of its premium assets, our example insurance company would no longer have enough funds to cover its expected claims. That is not acceptable and so the premium surplus would have to be rebuilt by raising new equity or debt capital, both of which could have a negative impact on shareholder wealth.

This is more or less what happened to many insurance companies such as RSA and Aviva after the dot-com bubble burst. It was a major problem for them and their shareholders.

However, if our example company had £100m in premium assets and £90m of claim liabilities then it would have a premium surplus of £10m. In that case if the premium assets declined in value by 2% to £98m the company would still have a sufficient capital buffer and no additional capital would need to be raised. Shareholders in this company would likely be much better off than in the previous example.

There are lots of different ways to measure an insurance company’s capital buffer, but I like to use a traditional measure of capital strength known as the premium to surplus ratio.

As the name suggests, the premium to surplus ratio is the ratio of premium income (net written premium) to premium surplus (tangible shareholder equity for the sake of simplicity). This ratio tells you whether an insurer is writing a sensible amount of new insurance given its existing premium surplus.

As a quick recap, here are some definitions:

Definitions

Net written premium – The amount of premium written during the period in question (typically the financial year) net of reinsurance premiums (where reinsurance is effectively insurance taken out by an insurance company to cover some portion of its claim liabilities). For example, a car insurance policy with annual premium of £120 written one month before the end of the financial year would represent £120 of premium written.

Net earned premium – Net earned premium is the figure reported on the income statement and as such may be easier to find than net written premium. For mature insurers the two values are usually very similar. This is the amount of premium an insurance company has earned during the period. Using the previous example of a £120 policy written exactly one month before the year end, the premium earned during the period would be for just that one month, i.e. £10, not £120.

Tangible shareholder equity – Shareholder equity minus any intangible assets and goodwill. This is the capital buffer which should ensure the company can fulfil its claim liabilities in good times and bad.

Calculating the average premium to surplus ratio

To calculate the five-year average premium to surplus ratio you’ll need the following figures covering the last five years:

1. Net written premium – This is not always quoted in the income statement so you’ll just have to search for it in the annual results (or use net earned premium from the income statement).
2. Total equity – On the balance sheet.
3. Intangible assets – Listed under non-current assets on the balance sheet.

This is a very simple calculation:

Calculating the average premium to surplus ratio

1) Calculate tangible equity:

tangible equity = total equity – intangible assets

2) Calculate the premium to surplus ratio:

premium to surplus ratio = net written premium / tangible equity

3) Calculate a five-year average for the premium to surplus ratio

DONE UP TO HERE

What is a prudent premium to surplus ratio?

From both research and experience I have found that a premium to surplus ratio of 2 seems to be a reasonable cut-off point for what could be considered a conservative premium surplus.

Beyond that level insurance companies may be more sensitive to negative shocks and surprises, potentially leading to dividend cuts and/ or rights issues. Companies whose ratio is consistently below that limit may be more robust when bad things happen, as they invariably will at some point.

Defensive value rule of thumb

Only invest in an insurance company if its five-year average premium to surplus ratio is less than 2.

Here is a worked example from an insurance company whose shares I have owned quite recently.

RSA Group’s premium to surplus ratio RSA (previously called Royal & Sun Alliance) is a FTSE 100-listed general insurance company, which means it primarily insures things like cars and houses. It has had a pretty rough time of it since the year 2000, with rights issues and dividend cuts on more than one occasion.

Table 4.6 shows how thick its capital buffer has been over the past five years.

TABLE 4.6

Table 4.6: RSA Group’s average premium to surplus ratio for the five years to 2014

Let’s work through the steps to calculate the average premium to surplus ratio:

1. Calculate tangible equity for each of the last five years (here’s the 2010 calculation):
   1. 2010 tangible equity = £3,895m – £1,209m = £2,686m
      
2. Calculate the premium to surplus ratio for each of the last five years (again, just for 2010):
   1. 2010 premium to surplus ratio = £ 7,455m / £ 2,686 = 2.8
      
3. Calculate a five-year average for the premium to surplus ratio:
   1. five-year average premium to surplus ratio = (2.8 + 3.2 + 3.5 + 4.5 + 2.4) / 5 = 3.3
      

RSA’s average premium to surplus ratio is 3.3, which is clearly too high for my liking. To make matters worse it is above my preferred maximum of 2 in every single year.

It is perhaps no coincidence that RSA suspended its final dividend for 2013 and launched a £ 775m rights issue in 2014. If it had maintained a significantly lower premium to surplus ratio then perhaps neither of those undesirable actions would have been required.

Amlin’s premium to surplus ratio

Let’s have another look at Amlin to see if it has a substantial capital buffer to go along with the high rates of profitability we uncovered earlier. You can see Amlin’s average premium to surplus ratio in Table 4.7.

TABLE 4.7

Table 4.7: Amlin’s average premium to surplus ratio for the five years to 2014 The premium to surplus ratio for Amlin is below 2 in every single year and therefore – of course – its five-year average is well below 2 as well. Perhaps this is one of the reasons why Amlin was able to maintain its dividend in 2011 even though the company made a substantial loss in that year after several major earthquakes resulted in exceptionally high levels of claims.

As with any financial measure, the premium to surplus ratio is not perfect and a high or low value does not automatically mean an insurance company will or won’t run into problems. However, on balance I think it’s a useful ratio and quite easy to calculate. It can, in many cases, separate out the riskier insurance companies from the not so risky.