1. Ally Financial
2. Citigroup (C)
3. SunTrust (STI)
4. MetLife (MET)
One of these is not like the others. MetLife is a life insurance company. Not a bank.
This raises a few questions:
1. Why was MetLife stress tested?
2. Why did MetLife fail its stress test?
3. Is the method used to stress test banks appropriate for a life insurer?
4. Does it matter that MetLife failed its stress test?
Let’s take the last question first. To MetLife shareholders it does matter. Because MetLIfe will not be able to raise its dividend and buy back billions of dollars of its stock as the company had hoped to do.
Does it matter to the country, the world, the global financial system, etc.?
I’ve always had mixed feeling on this point. You’ll remember that during the financial crisis some big insurers talked about getting capital. Now, one question is why an insurer – except to the extent they insure financial products – needs additional capital during a banking crisis.
After all, banks depend on deposits. That’s how they fund themselves. People can pull their deposits. Even very healthy banks can fail in a crisis if people stop trusting that bank.
But what about insurers?
Why would an insurer be at risk during a financial panic?
MetLife is very, very highly leveraged. Heading into the summer of 2008 MetLife – a company with a market cap fluctuating around $30 billion to $45 billion in the time it had been a public company – had total assets of $556 billion. The majority of these assets (61%) were investments. Back in the summer of 2008, MetLife’s biggest investment categories looked something like this:
· Bonds: $241 billion
· Loans (Consumer/Mortgage): $49 billion
· Loans (Policy): $11 billion
· Real Estate: $7 billion
· Stocks: $5 billion
And tangible shareholder’s equity was $27 billion.
So, investments relative to tangible equity were:
· Bonds: 9 times tangible equity
· Loans (Consumer/Mortgage): 2 times tangible equity
· Loans (Policy): 40% of tangible equity
· Real Estate: 25% of tangible equity
· Stocks: 20% of tangible equity
So, in the summer of 2008, if MetLife’s bonds lost 12% of their value – MetLife would lose 106% of its shareholder’s tangible equity.
If MetLife’s bonds lost 14% of their value, the company’s intangibles would have been wiped out as well. Leaving a negative number on the “total stockholders’ equity” line of the balance sheet.
What kind of bonds did MetLife own in the summer of 2008?
· Corporate: 48%
· Residential Mortgage: 23%
· Commercial Mortgage: 8%
· U.S. Treasuries: 8%
· Foreign Sovereigns: 6%
Since MetLife’s bond portfolio was about 9 times the size of its tangible equity, multiplying those percentages by9 will give you an idea of how big these investments were relative to MetLife’s tangible equity. As an example, a 50% loss on residential mortgage bonds would have wiped out all of MetLife’s tangible equity because residential mortgage bonds were $56 billion and MetLife only had $27 billion in tangible equity back in 2008. A loss of 25% on MetLife’s corporate bond portfolio would be enough to eliminate all of MetLife’s tangible equity.
But, that was back in 2008. What does MetLife’s balance sheet look like today?
|Bonds||$241 billion||$354 billion|
|Loans (Consumer/Mortgage)||$49 billion||$63 billion|
|Loans (Policy)||$11 billion||$12 billion|
|Real Estate||$7 billion||$8 billion|
|Stocks||$5 billion||$3 billion|
|Tangible Equity||$27 billion||$49 billion|
And let’s take a look at the five bond categories as a percent of MetLife’s overall bond portfolio back in 2008 and what those same categories account for today.
It’s not worth discussing individual issuers. MetLife’s investment portfolio is extremely diversified. The only thing worth mentioning is that about 40% of MetLife’s foreign government bonds are Japanese government bonds. MetLife owns more than $20 billion in Japanese government bonds.
Exposure to individual corporations is unimportant. MetLife only has $12 billion in the top ten names in its corporate bond portfolio. So, the risk at MetLife does not come from specific defaults – unless you’re worried about the American and Japanese governments defaulting – it comes from losses across all corporate bonds.
I can’t get into everything MetLife owns in this article. The company has asset backed securities. So they effectively own more than $4 billion in credit card loans, more than $2 billion in student loans, more than $1 billion in car loans, etc.
These do account for some of MetLife’s unrealized losses – but the bulk of MetLife’s unrealized losses in any financial crisis will come from its corporate bonds.
Okay. Let’s take a look at how bad it got for MetLife in the financial crisis.
The worst balance sheet the company ever filed with the SEC was its balance sheet for March 31st, 2009. The company had just $18 billion in tangible equity. And total assets of $491 billion. That suggests a leverage ratio of about 27. In reality, part of MetLife’s business involves carrying assets and liabilities that belong to separate accounts. When you adjust for this MetLife’s actual leverage ratio was about 21 to 1 in the middle of the crisis. Or to put this another way – MetLife got to the point where shareholders' equity less goodwill was only 5% of total liabilities.
They also included this disturbing note in their 10-Q:
“During the three months ended March 31, 2009, a period of market disruption, internal asset transfers were utilized extensively to preserve economic value for MetLife by transferring assets across business segments instead of selling them to external parties at depressed market prices. Securities with an estimated fair value of $3.7 billion were transferred across business segments in the three months ended March 31, 2009 generating $509 million in net investment losses, principally within Individual and Institutional, with the offset in Corporate & Other’s net investment gains (losses). Transfers of securities out of the securities lending portfolio to other investment portfolios in exchange for cash and short-term investments represented the majority of the internal asset transfers during this period.”
A lot of MetLife’s assets are not valued using market quotes alone. In fact, only the company’s U.S. Treasuries and stocks consistently fall into the Level 1 fair value measurement category
So, the fact that MetLife made extensive use of internal asset transfers during the 2008-2009 crisis is obviously a big concern. But it may say more about the state of financial markets than it does about the state of MetLife in early 2009.
So, that’s as bad as things got for MetLife back in 2009.
Now, let’s come back to today.
Today, MetLife released this statement:
“MetLife is financially strong and well positioned for both the current environment and a potential further economic downturn. We are deeply disappointed with the Federal Reserve’s announcement. We do not believe that the bank-centric methodologies used under the (Comprehensive Capital Analysis and Review) are appropriate for insurance companies, which operate under a different business model than banks…The established ratios used to measure insurance company capital adequacy, such as the (National Association of Insurance Commissioners) risk-based capital ratio, show that MetLife is financially strong. At year-end 2011, MetLife had a consolidated risk-based capital ratio of 450%, well in excess of regulatory minimums…. At year-end 2011, MetLife had excess capital of $3.5 billion. We project our excess capital will grow to $6 billion to $7 billion at year-end 2012, before any capital distribution actions. It continues to be our strong belief that excess capital should be returned to shareholders and we remain fully committed to doing so…In the capital plan we submitted, we requested approval for $2 billion in stock repurchases and an increase of MetLife’s annual common stock dividend from $0.74 per share to $1.10 per share….MetLife emerged from the 2008-2009 financial crisis in a position of strength, and was the only one of the top 19 bank holding companies that did not participate in the government’s Troubled Asset Relief Program. Today, we continue to demonstrate significant financial strength…”
So that’s the company line. Is it the truth?
Before we take a look at what the Federal Reserve has to say, let’s check in with A.M. Best. They rate insurers and they updated their outlook for MetLife back in November 2011:
…revised the outlook to stable from negative and affirmed the financial strength rating..of A+ (Superior) and issuer credit ratings…of "aa-" of the primary life/health insurance subsidiaries of MetLife… These ratings reflect MetLife's diverse business mix, prominent market position and brand recognition in several business lines, favorable operating results and significant operating scale. These strengths are enhanced by the significant financial flexibility, diverse sources of revenue and earnings and the expanding global market footprint of MetLife… Partially offsetting these positive rating factors is MetLife's overall risk appetite and risk-adjusted capital position…which is viewed as somewhat low for its current rating level. A.M. Best continues to have concerns regarding the company's high exposure to real estate linked assets, primarily from its large commercial mortgage loan portfolio and real estate holdings. A.M. Best believes MetLife's future earnings will be pressured as the low interest rate environment continues with additional strain on the company's interest sensitive product margins result in further spread compression. Recently, MetLife's exposure to asset accumulation products with embedded guarantees has grown significantly, which subjects the company to increased exposure to equity market volatility.”
What I did is somewhat unfair. A comparison of A.M. Best’s analysis of MetLife’s financial condition with the Fed’s stress test make the stress test look silly. A.M. Best correctly cites MetLife’s operating strengths but also mentions that MetLife has less capital than you’d like to see an insurer have and is taking new risks with its asset accumulation products. Long-term, if these risks are mismanaged the company could end up in a lot of trouble. In fact, these products may be the biggest company specific risk MetLife faces.
That was A.M. Best. Now, let’s look at the Federal Reserve’s take on MetLife.
Here is the famed stress test (PDF).
Basically, the Fed says that MetLife could – under stress conditions – report $10.8 billion in net losses over a two year period. Now, MetLife is expected to earn $9.6 billion before investment losses, so this actually assumes over $19 billion in investment losses. These losses are expected to come exclusively from two categories: realized and unrealized losses on available for sale and held to maturity securities.
In other words, the Fed says MetLife could lose $20 billion in its investment portfolio.
How likely is that?
MetLife had $30 billion of unrealized losses in its investment portfolio in March 2009.
But here’s the interesting thing about the Fed’s stress test – which backs up MetLife’s assertion that it’s a dumb idea to apply a bank stress test to insurers – the Fed sees MetLife producing 25% of all the investment losses of the 19 institutions.
What’s interesting about this is that MetLife accounts for less than 5% of the combined assets of these 19 institutions. So, the Fed is saying MetLife is expected to have more than 5 times as many investment losses per dollar of assets as the other 18 institutions.
This makes sense. MetLife is an insurer. The other institutions are banks. One of them has a big investment portfolio. The rest have big loan portfolios.
Parts of the Fed’s stress test scenario make a lot of sense. But one part is never explained.
Why does MetLife need to maintain the same level of common equity to total assets as a bank?
What exactly would happen to MetLife if it fell below this line?
See, other banks have counterparty risks. And they have deposits that they can’t restrict access to. MetLife doesn’t.
Also, if you look at the minimum stressed ratios – which are the lowest quarterly ratio assuming no capital action – MetLife and Morgan Stanley both are shown having 5.4% tier 1 common ratios at their low point. If MetLife goes through with its plans, it would drop to 5.1%. Morgan Stanley’s would not change.
The problem with that is the assumption that at a tier 1 common ratio of 5.1% MetLife has an equal or greater chance of failure as Morgan Stanley. It doesn’t. If MetLife had a tier 1 common ratio of 5.1% sometime in the next couple years and Morgan Stanley had a tier 1 common ratio of 5.4% at the same moment in time – Morgan Stanley will fail first.
The biggest problem with MetLife being included in the stress test is this line from the Fed’s report:
“(For the group)..the biggest sources of loss are losses on the accrual loan portfolios and trading and counterparty losses from the global financial market shock. Together, these two account for 85 percent of the $534 billion in projected losses for the 19 (bank holding companies).”
These categories account for 85% of losses for all 19 institutions but less than 5% of MetLife’s losses.
So, this raises the question of whether investors should ignore the Fed’s stress test when looking at MetLife shares.
You should read the stress test for yourself. But, in general, I think it’s an extremely complicated and extremely useless way of assessing MetLife’s ability to withstand stressful market conditions.
Instead, we can take the 2008-2009 financial panic as our guide.
How much could MetLife really lose in a stressful situation?
About $35 billion.
· 15% loss on corporate bonds: $25 billion
· 20% loss on commercial mortgage bonds: $4 billion
· 23% loss on asset backed securities: $ 3 billion
· 8% loss on residential mortgage bonds: $3 billion
Those are the percentage paper losses MetLife had on March 31 2009.
They also had paper gains. And if we entered a replay of the 2008 crisis, MetLife could presumably have gains like:
· $3 billion on government bonds
After all, money has to go somewhere. And MetLife’s corporate bonds and government bonds would likely move in different directions at the absolute height of the panic.
But we won’t assume any offsets like that.
Instead, we’ll just apply the $35 billion in losses. Again, that just takes MetLife’s portfolio today and applies the loss rates – 15% on corporate bonds, 20% on CMSs, 23% on ABSs, and 8% on RMSs – that MetLife experienced in early 2009.
Under those circumstances, MetLife would have $49 billion of tangible equity to absorb $35 billion in paper losses. The Fed assumes that MetLife would earn $8.6 billion before investment losses during the two-year crisis period the Fed projected. We can compare this to what MetLife actually earned in pre-tax profits excluding investments gains and losses. Last year that number was $6 billion.
Back in the crisis years – 2008 and 2009 – it was $3.3 billion and $3.4 billion respectively. So, that’s a good guess for what MetLife’s pre-tax earning power in the middle of a financial panic would be – before investment losses.
Overall, then we have $49 billion in tangible equity today. And we have $6.7 billion in operating income earned over a two-year period to absorb up to $35 billion in paper losses.
That’s a more realistic idea of what MetLife would look like in a financial panic than what you see in the Fed’s stress test. Since it’s based on actual data from the 2008-2009 crisis.
At the low point, this could leave MetLife with as little as 3.5% of its total assets – excluding separate accounts – in tangible equity.
And that’s before any talk of dividend increases and share buybacks.
But how realistic is this stress test scenario? And should investors apply it in their thinking about a company like MetLife?
I don’t think so.
I think there are two problems with doing that:
1. The whole exercise of stress testing financial companies – just after a panic – to see how they would fare in a second panic.
2. Applying the same test to a life insurance company as you apply to banks.
I don’t think either idea makes much sense. The least important time to stress test financial companies is after they’ve been through a panic and have lowered their leverage.
As for what investors should do regarding MetLife – it’s unclear whether the stress test will have much long-term importance for the company.
MetLife is expected to stop being a bank holding company later this year.
The stock is down 5% as a write this – at $37.50 a share. Book value is $45.50. MetLife’s average return on equity is only about 10%. So, if you buy the stock today you shouldn’t fool yourself into thinking you are paying any less than 8 times normal earnings.
In a low interest rate environment, you may even be paying quite a bit more than 8 times normal earnings.
Ask Geoff a Question about MetLife
Check out the Buffett/Munger Bargains Newsletter
Check out the Ben Graham: Net-Net Newsletter