Over the last few years, Standard Chartered has not turned out to be a good investment, with shareholders being hit by both a rights issue and a suspended dividend.

The root cause was the bank’s balance sheet, which was not able to withstand significant loan impairments caused by a delayed aftershock of the financial crisis.

Having sold my Standard Chartered shares just a few days ago, I think now is a good time to look back at what happened. However, rather than simply crying over spilt milk, I want to focus on why the milk was spilt in the first place and how I (and perhaps you) can try to avoid this sort of unpleasant situation in future.

But before I get into the details, here’s a quick snapshot of the results of this investment.

• Purchase price and date: 1,215p on 07/07/2014
• Sale price and date: 744p on 06/03/2017
• Holding period: 2 years 8 months
• Capital gain: -40%
• Dividend income: 7%
• Annualised gain: -15 %

Overview: “Only when the tide goes out do you discover who’s been swimming naked” – Warren Buffett

When Standard Chartered joined the UK Value Investor model portfolio it was one of the only UK-listed banks to have survived the initial phase of the global financial crisis completely unharmed.

That was mostly down to the fact that the majority of its business was and is conducted in Asia, Africa and the Middle East, rather than in the West where most of the causes and effects of the crisis were centred.

In fact, Standard Chartered didn’t just survive the financial crisis; it sailed through it almost without blinking. It quickly settled back into its previous pattern of almost monotonously regular double-digit growth, but that happy picture was not to last.

Things started to go wrong in 2013 when growth in the emerging markets where Standard Chartered operates began to slow down, partly as a long-delayed reaction to the financial crisis. This reduced corporate finance activity which hurt the company’s wholesale banking business, and there were more headwinds to come.

Increasing regulatory costs, persistently low interest rates and collapsing commodity prices all combined to create a “perfect storm” that stopped Standard Chartered’s growth in its tracks and sent the company into reverse.

Eventually, the dividend was cut and then suspended and a £3 billion rights issue was needed in order to strengthen the balance sheet.

Here’s what the last couple of years looked like from the point of view of Standard Chartered’s share price, including the points at which I bought and sold the shares:

Buying: A “safe haven” bank facing normal, cyclical headwinds

When I bought Standard Chartered in 2014 it had an almost perfect track record. Its 10-year growth rate was very high at more than 10% and its return on equity had averaged 13% over the same period.

Its balance sheet also appeared to be strong, not only relative to regulatory minimums but also relative to most other UK-listed banks.

The chart below shows how attractive that track record looked:

That track record looked no less impressive when compared to the FTSE 100’s record over the same period. On top of that, the usual valuation premium associated with high-quality, high-growth companies was nowhere to be seen.

Thanks to the emerging market slowdown Standard Chartered’s yield was above average, while its PE10 and PD10 ratios were also better than average, as highlighted in the table below:

By every measure, Standard Chartered appeared to be an above-average company trading at a below-average price, which is exactly what I’m always looking for.

Of course, this argument would only make sense if it was reasonable to assume that the company could maintain and grow its economic value – in other words, its net asset value, earnings and dividends – over the medium to long term.

There was a realistic chance that it wouldn’t. After all, investors were worried about the impact the emerging market slowdown would have on the bank’s earnings and, perhaps more importantly, on the value of its assets (i.e. the loans it had made to many thousands of businesses and individuals).

Having spent some time researching the potential impact, there appeared to be no strong consensus among analysts and commentators. Yes, there were risks and the company’s performance was likely to be weak for a period of perhaps several years, but no, it was not obvious that a dividend cut, dividend suspension or rights issue was just around the corner.

It was on that basis that I decided to invest the usual three to four percent of my personal portfolio and the UKVI model portfolio into this most impressive-looking of banks.

Holding: Events prove that the balance sheet was not as strong as I’d hoped

The investment got underway more or less as expected. The economic situation deteriorated, bad loans started to mount up and the share price fell. This initial period of negative results is entirely typical of most value investments so I was not particularly worried, even when the share price fell by around 25%.

From a low point at the start of 2015, the share price recovered as management focused on balance sheet strength, cutting costs and reducing risk. The dividend was maintained at the 2014 annual results (published in March 2015) and this appeared to be a fairly standard “hunker-down” phase, with a pinprick of light visible at the end of the tunnel. Here’s a comment from CEO Peter Sands:

“Trading conditions remain challenging and the actions we are taking to de-risk, cut costs and build capital are having an impact on near term performance.  However, underlying business volumes generally remain strong.  We remain confident in the strength of our franchise, the opportunities in our markets and in our ability to build returns to an attractive level in the medium term.”

However, the CEO’s confidence was ill-founded and just a few short months after that comment was made a new CEO appeared in the shape of Bill Winters, along with a new management team. Shortly after that, the 2015 interim results announced a major decline in profits and a 50% dividend cut, and soon after that, the company announced a £3 billion rights issue and then suspended the dividend.

At this point, I began to reassess my assumptions about what constitutes a strong bank balance sheet.

Measuring strength in a bank’s balance sheet

Like other companies, the balance sheet of a bank if made up of assets and claims on those assets, otherwise known as liabilities. The major assets of a bank are the loans it provides while the liabilities are money it borrows in order to fund those loans.

The liabilities can be split into two main types: debt and equity, where equity is often referred to as capital in the banking world. Debt comes mostly in the form of deposit accounts, while equity or capital is money owed to shareholders, including retained profits.

The critical thing to understand when it comes to measuring bank balance sheet strength is that capital acts as a buffer to protect depositors when many of the bank’s loans are not repaid in full.

As a simplified example, imagine a bank that has £95m in deposits and £5m of shareholder capital.

That £100m of liabilities is then loaned out as a series of loans (which are assets to the bank) totalling £100m. However, many of these borrowers do not repay their loans in full, so the value of those loans (the bank’s assets) falls by £10m to £90m. Since assets and liabilities must balance, the bank’s liabilities must also be reduced to £90m.

It is the role of capital to take the first and hopefully full hit of any declines in asset values, and in this case, the first £5m of the £10m asset write-down can be borne by shareholder capital, but after that first £5m there is no capital left.

Unfortunately, the bank’s depositors must now take the remaining £5m hit, leaving depositors out of pocket and the bank insolvent. At this point, a rights issue, nationalisation or the failure of the bank are the only routes available.

It is therefore imperative that banks have enough capital to absorb loan losses without serious risk of failure because bank failures can lead to systemic runs on banks, which of course most societies and their governments want to avoid.

This is why banking regulation makes a big deal out of capital ratios, which measure the ratio between the amount of risk taken on by a bank and the ability of its capital to absorb losses on those risks.

For a deeper and better explanation of bank balance sheets, have a look at Bank Capital and Liquidity (PDF) from the Bank of England.

The most prominent capital ratio is the Common Equity Tier 1 (CET1) ratio, an evolution of the previous Core Equity Tier 1 ratio. This is the ratio between a bank’s risk-weighted assets and its “highest quality” capital and it’s the ratio I originally used to conclude that Standard Chartered was well-capitalised.

The current regulation requires banks to have a CET1 ratio of between 4.5% and 9.5%, depending on various factors, so my initial assumption was that a ratio of 10% could be described as adequate.

This seemed reasonable as the major UK banks all had CET1 ratios of less than 10% during the pre-crisis banking boom (where – with hindsight – they obviously didn’t have enough capital to absorb losses on their loans) whereas by 2013 (after years of rights issues and much strenuous effort to strengthen their balance sheets) they all had CET1 ratios of more than 10%.

In fact, Standard Chartered was already targeting a CET1 ratio of 11% to 12% in 2014. This turned out not to be enough, so the target was moved upwards to 12% to 13% during 2015, a target which was subsequently achieved by suspending the dividend and raising £3 billion of additional capital through a rights issue.

Given these negative events, I decided about a year ago to up my minimum CET1 ratio requirement to 12%, or more specifically that it should have averaged more than 12% over the last five years.

Now that I’ve sold Standard Chartered at a loss I’ve decided to raise that requirement even further. Here’s my new CET1 rule of thumb:

• Only invest in a bank if its Common Equity Tier 1 (CET1) ratio has been above 12% in every one of the last five years

However, rather than simply relying on a slightly more demanding requirement for the CET1 ratio I have also decided to include two additional capital ratios in my bank analysis process.

The Leverage Ratio (or the assets-to-capital ratio)

The leverage ratio is very similar to the CET1 ratio in that it compares assets to capital. But in the version of the leverage ratio that I’ll be using, assets are not risk-weighted and capital is taken to be total capital rather than common equity tier 1 capital.

This leverage ratio is easy to calculate. It’s simply tangible assets (total assets minus intangible assets) divided by tangible capital (tangible assets minus total liabilities), which is slightly confusing because the CET1 ratio is capital divided by assets rather than assets divided by capital.

Here’s a chart showing the leverage ratios for many of the UK’s major banks. The general trend towards less leverage and stronger balance sheets is obvious.

In 2008 at the peak of the pre-crisis banking mania, Standard Chartered, HSBC, Lloyds, Barclays and the Royal Bank of Scotland had leverage ratios that were far higher than they are today. Lloyd’s in particular was way “off the chart” with a leverage ratio of more than 80. Fast forward to 2016 and after a decade of capital building those leverage ratios are now all below 20.

As you can see, by this measure Standard Chartered (in red) was for much of the period the least leveraged and best-capitalised bank out of this group of admittedly seriously over-leveraged and under-capitalised banks. So investors were right to think that Standard Chartered’s balance sheet was relatively robust, but being robust relative to a collection of fragile banks is not the same thing as being robust in an absolute sense.

Since I’m going to use the leverage ratio when analysing banks in future, I need to set an absolute (rather than relative) hurdle rate. Since the rights issue, Standard Chartered’s leverage ratio has been slightly below 15, so I’m going to use that as my hurdle rate.

In my opinion, 15 is probably close to the perfect leverage ratio for banks that are under stress (and as a value investor I’m typically investing in companies that are under a not insignificant amount of stress).

Why? Because when companies raise capital through a rights issue in order to strengthen their balance sheet, they tend to “kitchen sink” it. In other words, management works out how much extra capital the company actually needs and then they try to raise that much plus an additional significant safety buffer.

They do that because if they don’t raise enough capital the first time around and end up having to go back to shareholders to raise even more capital, they will definitely lose their jobs.

So here’s my leverage ratio rule of thumb:

• Only invest in a bank if its leverage ratio (tangible assets / tangible capital) has been below 15 in every one of the last five years

This currently rules out every UK bank, but that’s a price I’m willing to pay in the name of risk reduction.

The Gross Revenue Ratio (or the revenue-to-capital ratio)

In this capital ratio, the measure of risk is gross revenue. The idea is that if two banks have the same total value of tangible assets (a reasonable proxy for the value of their loans), the bank with higher-risk loans will generate more interest income, so measuring gross revenue is a simplistic way to differentiate between banks with higher and lower risk loans.

Again, the ratio is easy to calculate. It’s simply gross revenues (interest income plus other income) divided by tangible capital (tangible assets minus total liabilities), expressed as a percentage.

Here’s another chart, this time showing the revenue ratios for those same major UK banks. As before, the deleveraging trend is easy to see.

All of the banks generate far less revenue (i.e. take on far less risk) per pound of capital today than they did before the banking crisis. This is of course a very good thing, for both the banks and society in general.

Gross revenue for most of the big banks is now less than 50% of tangible shareholder capital and my initial thought was to use that 50% figure as the maximum allowable gross revenue ratio.

However, some smaller banks are able to generate much higher margins on their loans than the big banks, mostly because they provide a much more bespoke service to smaller companies that the big banks are not interested in.

Higher margins give these banks a gross revenue ratio of more than 50% because they charge more interest per loan, but these banks are nowhere near being overleveraged according to the leverage ratio.

So as a compromise, I’m going to set the maximum gross revenue ratio at 100%, i.e. gross revenues should not exceed tangible capital, and as usual, I want this to be true over the last five years. Hopefully, this rule will be tight enough to exclude reckless banks whilst being open enough to not rule out prudent but high-margin banks.

Here’s the rule of thumb:

• Only invest in a bank if its gross revenue ratio (total revenue / tangible capital) has been below 100% in every one of the last five years

Most of the big banks now pass this test, partly because they generate such low returns from their loans. But that’s okay because a) it would have ruled them out before and during the financial crisis and b) they still fail the leverage ratio rule.

If you want to know more about these ratios you could do worse than read Capital Ratios as Predictors of Bank Failure (PDF), from the Federal Reserve Bank of New York.

Selling: A suspended dividend is an automatic sell signal

As a result of Standard Chartered failing to pay a dividend for more than a full financial year, it has dropped out of the UKVI stock screen, which requires all companies to have an unbroken record of dividend payments going back at least ten years. This means Standard Chartered can no longer be compared with the other holdings in the model portfolio, or at least compared according to my investment strategy.

This – along with the fact that it is not at all obvious when the dividend will reappear – makes the bank an automatic sell, which is why I chose to sell it at the start of March, removing it from both my personal portfolio and the model portfolio.

The proceeds will, as usual, be reinvested next month into a hopefully more successful investment.