Measuring leverage in banks, insurance companies and non-financial companies

Companies use leverage because it can boost earnings, and that’s a good thing.

But leverage is also bad because it increases the volatility of earnings and increases the risk of bad things happening such as rights issues, dividend cuts and bankruptcy.

The trick is to find companies with the “right” amount of leverage given your investment goals, and that requires the right tools for measuring leverage.

Measuring leverage in non-financial companies

Leverage in non-financial companies is definitely easier to get to grips with than it is for financial companies like banks and insurance companies.

For the sort of relatively defensive companies I’m looking for, those with consistent records of profitable dividend growth, it basically comes down to borrowed money. How much has the company borrowed and how does that compare to its ability to pay the interest on those borrowings in both good times and bad?

There are lots of different ways to approach this. Two of the most popular are:

  • Interest cover, typically defined as operating profit divided by interest
  • Net debt to EBITDA, where net debt is total borrowings net of cash and EBITDA is earnings before interest, tax, depreciation and amortisation

Personally, I don’t use either of them because they rely on earnings from a single year, last year, which may not be a good guide to a company’s typical earnings over a number of years.

I think it makes more sense to compare debt or interest to a company’s average earnings over a number of years, in other words to its general “earnings power”. The result should be a ratio that is more reliable and robust than those that depend on the earnings of a single year.

As well as earnings over a number of years I prefer to use profit after tax in the ratio rather than EBIT or EBITDA because it has interest removed. That will increase the ratio (and decrease the allowable debt) for companies that have to pay a higher interest rate on borrowed money. So in some sense using profit after tax in the ratio combines some of the features of both debt and interest-based ratios.

Somewhat unimaginatively I call this ratio between borrowings and earnings power the Debt Ratio, and for clarity, it is defined as:

  • Debt Ratio = Total borrowings dividend by 10-year average profit (preferably adjusted profit) after tax

In my experience, a Debt Ratio of more than 5 is enough to put most companies into the danger zone, beyond which debt-related problems begin to be significantly more common. However, some companies may struggle with even less debt than that, so let’s take a quick look at why that might be the case.

Financing in cyclical and defensive industries

You can think of companies as operating in two types of industry, cyclical and defensive. Cyclical industries have industry cycles of boom and bust while defensive industries don’t, or at least to a smaller extent.

Companies in cyclical industries should generally hold less debt than companies in defensive industries, especially towards the peak of their industry cycle. It’s at the peak of a boom period when companies may have built up excess production capacity (more factories, equipment and so on) in order to meet peak demand.

When boom turns to bust the amount of money coming into the company may fall dramatically while the expense of any new factories and so on remains largely fixed.

This means the company may need to take on debt to fund those fixed overheads through the low part of the cycle. If it survives through to the next boom the company will earn handsome profits which it can use to pay down debts in preparation for the next downturn.

Because of these ups and downs and the unpredictability of earnings from cyclical companies relative to defensive companies, the maximum Debt Ratio that I would consider for companies in cyclical industries is 4 rather than 5. The idea is that this extra caution rules out riskier companies but isn’t so restrictive as to rule out lots of perfectly reasonable candidates.

Putting those points together produces the following two rules of thumb:

  • Don’t invest in a company that operates in a cyclical industry if its Debt Ratio is more than 4
  • Don’t invest in a company that operates in a defensive industry if its Debt Ratio is more than 5

For reference here’s a list of the official FTSE Sectors defined as either cyclical sectors or defensive sectors.

There are many other financial ratios that you could look at such as debt-to-equity ratios or working capital ratios, but in my experience, they aren’t so useful when applied to the sort of consistently successful companies I’m usually searching for.

You might think a single ratio like this would be overly simplistic, but having read up on research such as this Bank of England speech on the benefits of simple rules in complex and uncertain situations (aka The Dog and the Frisbee), I am quite happy to embrace simplicity.

As with any rule of thumb regarding which companies to invest in and which to avoid, there is a balance to be struck between excessive leniency and excessive caution. Of all the non-financial companies in the FTSE All-Share that have a 10-year unbroken dividend record:

  • 16 have no debt
  • 141 have some debt, with an average Debt Ratio of 2.4
  • 59 have a Debt Ratio of more than 5

Around three-quarters pass the test, which leaves a reasonably large pool of companies for further analysis.

Measuring leverage in banks

Banks are different to normal “trading” businesses. They don’t buy or sell services (or at least that isn’t their core business), instead, they borrow money from those who have it and then lend it on to those who need it.

Profits are primarily in the form of net interest income, i.e. the difference between interest income on loans and interest expenses on deposits. There’s a bit more to it than that of course, but that’s the basic picture.

So the entire business of a bank depends on borrowed deposits. The result is that most debt ratios (including the one above) are useless when analysing banks because they make banks look hopelessly over-leveraged.

What we need is a bank-specific leverage ratio which will help us to make reasonable assumptions about a bank’s underlying capital strength.

Fortunately, it seems that banking regulators have caught onto this idea as the new Basal III regulations stipulate, for the first time, a minimum leverage ratio for banks.

This leverage ratio is the ratio between the bank’s capital and its exposure to risk. An important ratio has long been the “Tier 1” ratio, which comes in a variety of guises, the strictest of which is Core Tier 1, based on Core Tier 1 capital.

That’s a pretty meaningless phrase to most people (including me before I started looking into this) but Core Tier 1 capital is basically a bank’s highest quality assets which can used as a balance against lending and other risky banking activities.

Here are the Core Tier 1 ratios of some well-known banks:

  • Barclays: 2007 = 7.8%, 2013 = 9.9%
  • HBOS: 2007 = 7.4%, 2008 = 6% (subsequently “rescued” by a merger with Lloyds)
  • HSBC: 2007 = 8.1%, 2013 = 12%
  • Lloyds: 2007 = 8%, 2014 = 11%
  • Standard Chartered: 2007 = 9.8%, 2013 = 11.8%

As you can see, in the Wild West days before the financial crisis most banks had lower Core Tier 1 ratios (or equivalents) than they do today.

But it’s hard to see much of a difference between the banks that had major problems in 2007/8 with those that didn’t. What was more important than leverage in the financial crisis was the kind of risks a bank was taking, and that’s the sort of thing a simple ratio can’t show.

However, this leverage ratio can still be useful if we look at its levels from before and after the crisis.

Before the crisis most banks had Core Tier 1 ratios below 10%; in other words, they had less than 10p of Core Tier 1 assets for every 100p of risk exposure. After the crisis, most have moved to ratios above 10%.

Of course, they’re being forced to increase capital and lower leverage because of tighter regulation, but regulations don’t last forever, as we saw when the deeply useful Glass-Steagall Act was repealed in the late 90s, leading directly to the financial crisis.

So just in case the banks do decide to embark on another debt-fuelled feeding frenzy at some point in the future, I have started using the following rule of thumb for banks:

  • Don’t invest in a bank where its Core Tier 1 ratio is below 10%

That could result in banks being off-limits as an investment for years or even decades at a time, but as the financial crisis showed, if can be very painful to be invested in banks at the wrong time.

And if a bank does make it past the test it will be very conservatively financed relative to recent history.

Measuring leverage in insurance companies

Unlike most banks, insurance companies do have a simple entry on the balance sheet called “borrowings”, just like non-financial companies. This means the borrowings to earnings power ratio above can be used for insurance companies too.

However, there is another form of leverage that we should also be interested in.

Insurance companies work by pooling small amounts of money from lots of policyholders in order to pay out infrequent but large claims when they arise. This means they’re usually sitting on a large pot of collected premiums that don’t belong to them. This is often referred to as insurance “float”.

This float needs to be enough so that all claims can be paid on time, in other words, there needs to be a surplus of premiums over expected claims. If there isn’t enough of a surplus then an insurance company may need to raise additional capital either through a bond issue or rights issue, neither of which is good news for shareholders.

So with insurance companies, we want to be sure that the company has a more than adequate amount of surplus capital relative to the scale of its business.

One traditional measure for this is the premium to surplus ratio, which I’ll define here as the ratio of net written premium to shareholder equity (not including debt capital). You can find both of those in the company’s accounts.

To get a feel for what an appropriate ratio might be, here is the premium-to-surplus ratio for some leading insurance companies:

  • Admiral: 2011 = 1.4, 2013 = 1.1
  • Amlin: 2010 = 1.1, 2013 = 1.4
  • Aviva: 2000 = 2.1, 2011 = 3.5
  • RSA: 2000 = 2.3, 2011 = 3.2

I chose the year 2000 as the date point for RSA and Aviva because both companies had a lot of problems shortly after the start of the millennium. Those problems were caused in part because they both had relatively weak capital positions compared to the amount of new business they were writing.

RSA and Aviva both had a premium-to-surplus ratio of more than 2 when they were hit by falling investment values after the dot-com boom. That weakened their capital position further (as their floats were partially invested in equities) which led to a major rights issue for RSA and a significant bond issue for Aviva.

More recently these two companies have run into problems again, and largely for the same reasons.

In 2011 Aviva and RSA still had premium to surplus ratios above 2 and shortly afterwards Aviva cut its dividend while RSA launched yet another rights issue.

Amlin and Admiral, on the other hand, have always had a premium-to-surplus ratio of less than 2, and both have been relatively free of the problems suffered by their more highly leveraged cousins (although Amin did suspend its dividend for a year in 2001, that was due to losses from the 9/11 World Trade Center attack).

A premium-to-surplus ratio of 2 is an interesting cut-off point as several old insurance industry books mention it as a measure of conservative underwriting, and on the evidence above I’m happy to use the following rule of thumb:

  • Don’t invest in an insurance company if its net written premium is more than twice its shareholder equity (excluding debt capital)

But there is another factor involved here too.

The strength of an insurance company’s capital base is dependent on how much surplus insurance float it has beyond its expected claims. If the float is invested in risky, volatile assets then a surplus which looks fine this year can become dangerously thin if the risky assets fall in value.

As I mentioned above, this was another problem that RSA and Aviva suffered from in the early 2000s.

For years the insurance industry had lucked into massive investment profits from rising stock markets as their investment portfolios were allocated increasingly to equities.

In 2000, at the height of the dot-com bubble, many insurance companies, including RSA and Aviva, had too much of their float invested in overvalued risky assets.

The subsequent collapse of the equity markets from 2000 to 2003 increased RSA’s and Aviva’s already high leverage ratios to unsustainable levels, which led to large bond issues, major rights issues and dividend cuts.

Today the picture is very different. Comparing their equity investment allocation then and now:

  • Admiral: 2000 (wasn’t listed yet), 2013 = 0%
  • Amin: 2000 = 17%, 2013 = 8.5%
  • Aviva: 2000 = 30%, 2013 = 1%
  • RSA: 2000 = 26%, 2013 = 4%

Insurance companies are far more cautious today than they were a decade or so ago. However, there are no guarantees that they won’t slip back into bad habits if we have another long bull market in equities, so I’ll be using the following rule of thumb:

  • Don’t invest in an insurance company if it has more than 10% of its investment capital in equities

As with the banking rule of thumb, this may rule out investments in the insurance industry for long periods, but nobody ever went bust by not investing in insurance companies. What’s more important is that only conservatively financed insurance companies make it into a defensive portfolio.

Finally, there is another piece to the puzzle of insurance companies. It doesn’t directly relate to leverage, but it is important so I’ll mention it here anyway.

As we’ve just seen, during the long boom in equity markets during the 1990s insurance companies were able to make huge amounts of money by investing their float in stock markets around the world.

In fact, they could make so much money through investments that they were willing to lose money on the underwriting side of their business just to get their hands on additional premiums to invest.

For example, if an insurance company writes policies with say £100m in premiums, but expects to have to pay out £105m in claims and £5m in expenses, it has made an underwriting loss of £10m and has a combined ratio (combination of claim and expense ratios) of 110%, where any ratio above 100% represents an underwriting loss.

However, if the company can generate a 20% investment return on that £100m then that would more than offset the underwriting losses and create a net profit.

That’s basically what a lot of insurance companies were doing during the go-go days of the dot-com boom.

Here are some combined ratios for the same insurance companies we looked at a moment ago:

  • Admiral: 2000 (not listed yet), 2013 = 89%
  • Amlin: 2000 = 111%, 2013 = 86%
  • Aviva: 2000 = 109%, 2013 = 97%
  • RSA: 2000 = 110%, 2013 = 97%

During the late 90s stock market bubble RSA and Aviva were losing money on almost every policy they wrote, just so they could build up a larger float to invest at what they thought were going to be high rates of return.

Unfortunately for their shareholders, it didn’t work out that way.

They didn’t have enough surplus capital to support the amount of premiums they were writing, and when the stock market bubble burst both companies saw their too-small capital bases shrink even further, which led to rights issues, dividend cuts and so on.

Amlin escaped that fate, but cut its equity exposure in September 2011 shortly after the 9/11 attack and focused instead on making money the old-fashioned way, through sound and profitable underwriting.

So here is another simple rule of thumb to avoid insurance companies that are losing money on their core underwriting business:

  • Don’t invest in an insurance company if its five-year average combined ratio is more than 95%

This rule only applies to non-life insurers because life insurers don’t use the combined ratio. 

In summary, banks and insurance companies can be horribly complex things to understand. However, the combination of these simple rules of thumb may help to reduce the time it takes to analyse them and ensure that only the most prudent financial and non-financial companies make it into a defensive portfolio.

Disclosure: I own shares in Standard Chartered, RSA, Amlin and Admiral (and I sold my Aviva shares at the start of 2014)

Author: John Kingham

I cover both the theory and practice of investing in high-quality UK dividend stocks for long-term income and growth.

12 thoughts on “Measuring leverage in banks, insurance companies and non-financial companies”

  1. Dear John, allow me one comment:

    “That’s a pretty meaningless phrase to most people (including me before I started looking into this) but Core Tier 1 capital is basically a bank’s highest quality assets which can used as a balance against lending and other risky banking activities.”

    This is not correct, The ratio says nthing about the quality of assets. Its a very artificial numer, dividing a certain definition of equity (including some weird stuff) dividided by an artificial “risk weighted” asset base. Personally, I would use for banks the ratio shareholders equity divided by total balance sheet amount.

    1. Hi

      I’m certainly no banking expert, so it’s difficult for me to take a particularly strong position given my weak knowledge on the subject. However, quoting Basal III directly ( ), paragraph 48:

      ” A key element of the new definition of capital is the greater focus on common equity, the highest quality component of a bank’s capital. ”

      So I’m using the phrase “quality” in the same context as Basal III.

      As for risk weighting my assumption here would be that a core tier 1 ratio of
      over 10% is prudent enough that risk weighting shouldn’t be an issue. However, Basal III is also introducing a core tier 1 non-risk weighted ratio which would probably be another reasonable thing to look at once it’s fully up and running.

      1. John,

        “Capital” is a balance sheet item on the liability side. So “quality capital” does not mean assets. A bank can have 100% crappy assets but stil “quality equity capital”.

        Again, as long as Italian Government bonds have zero risk weightings (–study.html) I would strongly argue that any ratios on that basis are absolutely meaningless.


  2. An interesting article. I work in a large bank and the Core Tier 1 ratio is the single most talked about metric. Pre crisis nobody ever talked about capital, it simply wasn’t an issue, whereas now every new transaction or restructure that is done only happens after serious consideration to the impact on Risk Weighted Assets (RWA). Capital is and will continue to be in short supply for banks for the foreseeable.

    Of course as with all regulation, the unintended consequences of this squeeze on capital are huge. Some of the rules the regulators stipulate to determine how to calculate your RWA often render the numbers severely disconnected from the underlying risks! However the focus on Tier 1 means we regularly turn away profitable business because the capital demand it too high. This is the new reality for banks.

    1. Hi UTMT, if you look back at bank annual reports pre-2007 then capital just doesn’t seem to be important. I guess that’s fair enough because when was the last banking crisis before 2007? A very very long time ago.

      Now of course as you say it’s front and centre both internally and externally. And a good thing too, although I’m waiting to see what the non-risk weighted leverage ratio looks like.

      As for capital being in “short supply”, that’s true relative to the capital binge we’ve had for a very long time, so perhaps that’s no bad thing.

    1. Hi Andrew, thanks for the links.

      Interesting that the first one uses the same 10% Core Tier 1 ratio as I do. As for valuation I don’t use price to tangible book for banks but instead use the same combination of factors to balance growth, quality and value that I use with non-financial companies.

      The second link mentions the bank efficiency ratio, where I guess my version of that is ROE, which I measure as the median over 10 years. I didn’t mention ROE in the post because it doesn’t directly measure leverage, although of course it is impacted by leverage.

  3. ‘Under the money tree’ makes a good argument for the RWA. Good business is refused lending. Most affected are ‘value’ companies and small cap.

    By the way it is a good article. I do not invest in banks, my general opinion is their future is pretty bleak for the next 10 years. If you research the Nikkei index, the worst performer was the financial sector in Japan. I expect this to repeat in the UK.
    In the financial sector I only invest in Provident Financial and recently bought back in Hiscox and just about it. I forgot, I have an Australian bank and financial institution as well.

    1. Hi Eugen, thanks. I think I agree that banks will face a tough future for a while yet. The only one I own is Standard Chartered, which is slightly different to the UK ones because of its very different geographical exposure, although it too is going through a rough patch at the moment.

      1. What about Aviva? It seems that its consolidation with Friends Life is more like rights issue in disguise.

        To my shame, I did not realise that Friends has such s good free cash flow. Obviously the new changes in the pension rules will put it under threat, but Aviva seems to believe it will be able to protect it and ‘switch’ the pensions to their wrap based investments.

        My issues with these life insurance is two fold: they still have their main business in the UK, with business models under threat; regulation tend to change all over the world and ban commission on selling insured based savings plans.

        What I cannot comprehend is the positive effect of AE and bulk annuities. AE at this monent has a negative effect on the bottom line (it costs them lots to setup the schemes and cannot book any revenue as there is no initial fee or charge), but it could be a positive for the long term as they build funds under management at an average of 0.65% per annum. Bulk annuities have a positive effect on the short term, as insurance companies usually book 5% revenue immediately, but with people living longer and probably lower investment return it has a negative effect overall. It also creates the risk of ‘write offs’ and also high risk for increase in the capital adequacy.

        I see more reasons to stat away and invest in IT or healthcare.

Comments are closed.