Sunday, April 15, 2007

Insurance Business Part 1

What is insurance?

Insurance is a contract which transfers risk from the customer (the insured) to the insurance company (the insurer). If, for example, an insured customer gets into a car accident, his insurance company ends up paying the bill. In return for undertaking this risk, the customer must pay the insurance company a premium upfront.

What are premiums?

Premiums for insurance companies are essentially equivalent to sales for retail companies. Insurance companies take in premiums from customers, from which they pay out losses and cover expenses. Written premiums refer to the amount of new business an insurance company writes or “sells” to customers each year. If an automobile insurance company acquires 100 new customers in a year, with each contract requiring customers to pay $1000 in premiums, then that insurance company’s written premiums is $1 million for that year.

So far, it is pretty straightforward, isn’t it? The total amount of premiums an insurer is entitled to receive from its customers over the life of their insurance contracts is the gross written premiums.

What are net premiums?

Risks come in all kinds of shapes and sizes, and at times, an insurance company does not want to take on certain risks, or it wants to transfer some of its risk to another insurer (known as reinsurer). The insurance company must pay reinsurance premiums to the reinsurer. This reinsurance cost must be subtracted from the gross premiums, and the result is the net premiums. Just as net sales are a better measure of a retail business, it is the same for an insurance business.

What are net premiums earned?

IN GAAP, accrual-based accounting states that revenues and costs must be matched to the periods for which they are applicable. In other words, if a customer pays you today for a service to be rendered in a year, you cannot recognize that revenue and the associated costs until the service is performed. Similarly, since insurance contracts are often written for multi-year periods, the portion of the premium earned must be recognized on an accrual basis. For instance, if a customer pays an insurer $10 million in premiums in order to insure its risk for 10 years, then every year it is earning a tenth of the total net premium written, so its yearly net premiums earned is $1 million.

In reality, like any other type of company, the more business an insurer does, the better (assuming the business is writing the policy with the associated risks in the correct manner, of course). The companies that are able to generate a lot of premiums are generally more valuable. Some investors like to use price-to-sales ratios (market cap divided by sales) when judging retail business. In a way, it’s to a certain extent worth taking a glance at the price-to-net-premiums-earned ratio in order to ascertain an insurance company’s premium generating ability. Not forgetting that insurers have other sources of revenue besides premiums because premiums by itself is often insufficient to cover all associated losses and expenses if it is not put to work. And such revenue that flows through to the net income of the insurance company varies widely depending on the company.

Some of the top insurance companies are White Mountains, USAA, Mercury General, Progressive, Geico, General Re, W.R. Berkley, and Merkel.

Like great value investors who only load up heavily when the bet is highly mispriced, great insurance companies do business only when risk-adjusted returns are favorable. As such, investors should look for insurance companies that grow net premiums earning when risk premiums are high which usually occurs after a catastrophe or an industry shake-out, and stay disciplined when risk premiums are low. Interestingly, this is the simple formula Warren Buffett has stuck to turn Berkshire Hathaway which owns Geico and General Re, into the insurance juggernaut it is today.

What is float?

In the insurance industry, premiums held or written are other people’s money (OPM) till such time when the liability for the contract is over and done with. OPM is also known as float. And float is the heart of the insurance business that determines the moat of the company. Float has its cost or conversely has a value to it depending on how the float is deployed between the day the insurance receive the premium to the day it has to pay out all costs associated with the written premiums. Float is wonderful if it has no cost or if it doesn’t come at a high price, and over and above this, if it produces a cost associated with the float which is less than the cost the company would otherwise incur to obtain the funds.

Warren Buffett once remarked that float “has cost us nothing, and in fact has made us money.” Therein lay an accounting irony: “Through our float is shown on our balance sheet as a liability, it has had a value to Berkshire greater than an equal amount of net worth would have had.”

When you pay the premium on your automobile insurance, those premiums help cover operating expenses and go toward paying automobile claims from customers who get into accidents. The great thing about premiums is that insurer collects the money upfront but doesn’t have to pay out claims until much later down the road. In the meantime, the company “floats” these unpaid premiums by investing in stocks, bonds, and other securities to pocket a profit.

Float is really a valuable form of capital because not only does the insurance company get to keep the investment income, but also the company’s cost of capital is often low or even positive.
Here’s is how float is calculated: Net loss reserves + loss adjustment expense + funds held under reinsurance assumed + unearned premium reserves – insurance related receivables – prepaid acquisition costs – prepaid taxes – deferred charges applicable to assumed reinsurance.

Wow, that is really a handful but fortunately as always, it can be simplified to say that float is simply cash received from the customers that have not been paid out yet for claims and expenses.

The more float a company has, the more investment income it can generate. That’s a wonderful concept which Warren brought to Berkshire and the resulting outcome is Berkshire’s float increased from $20 million to $49 billion in the period from 1967 to 2005.

What is short tail and long tail claims?

The length of time between receiving a premium and eventually paying out claims affect how profitable float can be. The longer it takes to pay out claims, the more value the float can creates because ultimately what value a float can create is dependent on the basic notion of investing – you need time to create value, and many times, patience for a business to realize any value which is undervalued. For instance, Progressive and Mercury General write automobile insurance that is generally paid out quickly. If an insured driver gets in an accident, then the claim for vehicle damage and bodily injury is paid out soon. This is known as short tail insurance. On the other hand, insurance companies that insure liabilities like asbestos claims take a long time between getting the premium on hand to the day it is paid out because of litigation that remains for years and years. In general, such long tail insurance is preferable because the float can be reinvested over longer periods of time and compounding it.

What is the importance of loss reserve in the business of insurance?

As described earlier, accrual accounting states that revenues should be recognized when earned and that expenses should be matched to revenues in the same period. Therefore, insurers have to recognize their losses in the same period as they earn their premiums. The tricky part is insurers often do not pay claims until some future date, and this requires them to estimate their losses and creates a reserve for this estimate for future claims.

Estimating claims losses is sort of like trying to guess how drunk you are after each successive shot of tequila. Since the drunkenness doesn’t come until later, you have to make educated guesses based on statistics such a body weight and historical experiences. Insurers estimate their losses using actuarial estimates and experiences too.

If, in the aforementioned situation, you’re too conservative, you won’t have enough to drink and won’t achieve your desired state of intoxication. Similarly, insurers who are overly conservative may allocate too much of their capital to reserves and be underleveraged which depressed their investment income.

Conversely, if you’re too aggressive and have too much tequila, then you may end up your night prematurely due to excessive drunkenness, and you may end up with a hangover or even alcohol poisoning. Similarly, insurers who are too aggressive with their loss reserves may find that later on, they have to recognize losses and may even become insolvent.

One thing is clear though between the selection from these two approaches – it’s better to err on the side of caution just like all value investors’ approach to investing with caution. Even if an insurer can get away with aggressive loss reserves in the short run, claims must eventually be paid, and overly aggressive insurers end up paying the piper more often than not.

If loss reserves are management estimates, then how do investors get comfortable investing their hard-earned money in an insurance company? Reputation and historical track record are some ways to gauge an insurer’s creditability. If I’m at a golf course with Tiger Woods and my neighbour, Billy, and both of them tell me they have awesome handicaps, I’d tend to put more faith in Tiger Woods. Likewise, insurance managers who have great reputations mostly are due to long track records of conservative approach and also accurately estimating loss reserves.

As time passes and claims get paid, loss estimates become much more accurate. Every year, as new information unfolds, the insurer must re-estimate historical loss reserves, resulting in a loss adjustment expense. Overly aggressive insurers who under-reserve their losses and overstate income, eventually realized their claims are higher than their reserves, and have to make an upward loss adjustment expenses.

Insurers report their historical results in cumulative redundancy tables. If loss reserves were too aggressive, then the results show a large deficiency. If they were too conservative, then the result shows a large redundancy.

Practices of insurance all-stars

Some insurance companies have demonstrated over a stretch of years and underwriting cycles that they have the culture and integrity to adequately measure loss reserves. Investors would be well-served to stick to management teams that have already proven their mettle. As the saying goes, “If you don’t know the horse, know the jockey.” Some insurance companies with rock-solid “jockeys” are Geico, General Re, Berkshire Hathaway, Markel, Cincinnati Financial, and U.S.A.A.

How to evaluate an insurance company?

The key determinants are 1) the amount of float that the business generates, 2) its cost, and 3) the long term outlook for both these factors which is the most critical.

To start, float arises because premiums are received before losses are paid out with an interval that sometimes extends over many years depending on the nature of the risk insured – long tail or short tail. During the interval, the float is invested in securities. This pleasant event typically carries with it a downside: The premiums that an insurer takes in usually do not cover the losses and expenses it must eventually pay out. An underwriting loss would occur in this case, which is the cost of the float.

Thus, an insurance business only has value if its cost of float over time is less than the cost the company would otherwise incur to obtain the funds. But the business is a lemon if its cost of float is higher than market rate of money.

Because loss costs must be estimated, insurers have enormous latitude in figuring their underwriting results – there are two extremes, conservative or aggressive – and that makes it very tough for investors to calculate a company’s true cost of float. Errors of estimation, usually innocent but sometimes not, can be huge. The consequences of these miscalculations will flow directly to earnings. An experienced observer can usually detect large-scale errors in reserves, but the general public could not and at times, even these numbers are implicitly blessed by big names auditors. Thus, both the income statement and balance sheet can be a minefield in insurance business if the loss estimation is horribly overdone.

Hence, it pays to strive to be both consistent and conservative in reserving of loss. Occasionally, mistakes will be made and it is inevitable. And a warning here is there is nothing symmetrical about surprises in the world of insurance – they’re almost invariably unpleasant.

In insurance, an underwriting profit means there is zero cost of float. In other words, you get paid for holding other people’s money. If an underwriting loss occurs, it does not means the business is a lemon, it is a lemon only if the cost of float is more than the cost the company would otherwise obtain through other means.

Why is long tail better than short tail insurance?

Simply, the odds are better and time is on your side. Some years back, a few insurers that were then experiencing large losses offloaded a significant portion of these to Berkshire Hathaway which penalized its then current earnings but it gave them float they can use for many years to come. After the loss that they incur in the first year of the policy, there’re no further costs attached to this business.

If these policies are properly priced, the pain-today, gain-tomorrow effects should be taken. These transactions will materially distort the current earnings but provide great long term favorable results. Many reinsurers have little taste for this insurance. They simply can’t stomach what huge underwriting losses can do to their reported results, even though these losses are produced by policies whose overall economics are certain to be favorable. Therefore, investors should be careful in comparing underwriting results of different insurers.

An even more extreme long tail kind of insurance which you won’t find much of elsewhere besides at Berkshire arises from transactions in which they assume past losses of a company that wants to put its trouble behind it. To illustrate, the ABC insurance company might have last year bought a policy obligating Berkshire to pay the first $1 billion of losses and loss adjustment expenses from events that happened in, say, 1995 and earlier years. These contracts can be very large though a cap must be in placed to limit the exposure.

Under GAAP accounting, this “retroactive” insurance neither benefits nor penalizes the current earnings. Instead, an asset called “deferred charges applicable to assumed insurance” is set up. This amount reflects the difference between the premium received and the losses that it is expected to be paid (for which reserves are immediately established). This asset is then amortized by making annual charges to earnings that create equivalent underwriting losses. By their nature, the losses will continue for many years, often stretching for decades. As an offset, they have the use of float, lots of it.

Clearly, float carrying an annual cost of this kind is not as desirable as float that is generated from policies that are expected to produce an underwriting profit. Nevertheless, this retroactive insurance is still a decent business though it is not the better of the two.

And the ultimate benefit that is derived from float will depend not only on its cost but fully as important, how effectively it is deployed.

An example of long tail and low probability insurance is best found in transactions done at Berkshire. Ajit Jain is the all-star underwriting insurer manager in the world of insurance. Single-handedly, many deals were consummated due to his brilliance. An example is he negotiated a very interesting deal with, an internet company whose goal is to attract millions of people to its site and then to extract information from them that would be useful to marketers. To lure these people, held out the possibility of a $1 billion prize (having a present value of $170 million) and Berkshire insured its payment. A message on the site explained that the chance of anyone winning the prize is minute, and indeed no one won. But the possibility of a win is far from zero.

Writing such policy, a modest premium is received, face the possibility of a huge loss and get good odds. Very few insurers like that equation. Because each policy has its unusual and sometimes unique characteristics, insurers can’t lay off the occasional shock loss through their standard reinsurance arrangements. Thus, any insurance CEO doing a business like this must run the small, but real risk of a horrible quarterly earnings number, one that he would not enjoy explaining to his board or shareholders. However, any proposition that makes compelling mathematical sense regardless of its effect on quarterly earnings should be embraced for long term gain.

No comments: