Monthly Archives: December 2012

RMS: V11 made us proud

RMS takes a victory lap for the performance of much maligned v11 after the model’s performance on Sandy:

However, after Sandy, RMS says its evolved
model has proven to be more accurate in
estimating the impacts of wind, and
especially storm surge—the dominant
cause of losses related to Sandy…

The company points particularly to its modeling of storm surge. The PC360 article is a bit vague, but it sounds like old models keyed storm surge primarily off wind speed. V11 looks at geographic formations street by street.

Update: Fixed link above.

Reserving regulations

David Merkel blasts away at principles-based reserving:

You don’t want to hand over reserving rules to one hired by the company, no matter how ethical he might be.  That way lies disaster.  There are always subtle pressures put on actuaries to be less conservative, because companies face pressure to show good earnings in the short-run.

Think of the mostly European quants, accountants and actuaries using the Basel standards.  Giving them the authority to set their own reserves for credit using internal models led to setting the reserves too low.  You want to have checks and balances.  You don’t want to have players serving as their own umpires.  So what if the statutory standards are too tight?  That just means earnings will be delayed, not eliminated.  Risk margins should be received as earned, and never capitalized.  Besides, the current crisis shows us that we never truly understand the parameters of the distribution.

He’s reacting to this (sub. reg.) Wall Street Journal piece:

State regulators Sunday approved new rules for how U.S. life insurers set reserves for future claims, a decision that may ultimately free up billions of dollars for acquisitions, stock buybacks, dividend increases and other uses designed to boost the industry’s flagging returns.

I’ll put my potential conflict of interest up front – I do work occasionally with state insurance departments.

This idea has been kicking around the NAIC for a few years. Insurance commissioners are working trying to move away from prescriptive regulation – setting down a bunch of rules and making sure companies follow them. They are trying to move toward principles-based regulation – setting a goal for companies to strive toward.

The distinction has eluded me at time, in part because the words prescriptive and principles sound alike. But also, I’m a casualty actuary, and we have always had a principles-based approach to reserve-setting. P/C companies create their own models, then share them with regulators.

Now, property casualty contracts are probably easier to model than life contracts. Most p/c contracts expire without an obligation (no claim to pay). I’ve never worked on the life side, but I gather virtually all life contracts involve a payment at the end, and estimating the percentage that lapse is itself an important calculation that rises or falls with the economy. Add to that the interest rate and mortality assumptions embedded in the contract, and I can see David’s concern.

He may be a bit too concerned, though. The impairment rate on p/c companies is about 1%, and most insolvencies are so tiny they barely make the trade press. And regulators aren’t the only ones who will review reserving models. So will external auditors and ratings agencies. And they review them more frequently than regulators do and, often, with greater rigor.


About 1% of P/C companies become impaired in a year.


Small business and Obamacare

per state premiumsHere’s a big New York Times story about the dilemma small business owners face – should they offer health insurance to their employees? The article focuses on restaurants and hospitality workers, many of whom are young part-timers.

It’s a tough question, and lousy reporting makes it tougher. The article itself is Exhibit A. It exaggerates the costs and omits the subsidies many small business owners for which small business owners are eligible. Towit:

Employee health coverage now averages nearly $6,000 for an individual plan. That is considerable for businesses like restaurants in which the majority of workers make $24,000 a year or less, according to research by the Kaiser Family Foundation.

The article sure makes it sound like employee costs will rise 25%.

I have no idea where the Times got that $6K per person estimate. Kaiser Family Foundation, which is probably the most honest broker of Obamacare information, reports that the average individual plan cost $215 per month in 2010. The handy map you are looking at shows how costs vary by state.

Even factoring in health care inflation, Kaiser’s estimate is half what the Times reported. The most expensive state, Massachusetts, had average premiums of $5,244 per year, or 13% less than what the Times reported.

I pay about $6,800 per year, incidentally, and live in New Jersey, where premiums are 50% above the national average.

And I’m 52 years old, and that’s significant in this discussion. My premiums are triple what a 20-year-old would pay (maybe more, I lost the rate card my insurer sent), and of course your typical wait staff is closer to 20 than to 50.

On top of all that, my health plan is fairly rich – out-of-pocket costs can’t exceed $5,000. There are cheaper plans.

In fact, to get an idea, I browsed through, a government web site that compares health plans. For a 20-year-old, nonsmoking male in Essex Count, the cheapest plan was $150 a month, or $1,800 per year. It’s not a great plan – it covers zero out of network costs. On the other hand, it’s individual coverage, which tends to cost more than group plans.

And of course, employers doesn’t have to pay the entire premium. But Obamacare subsidizes the cost of what they do pay. The feds will reimburse via the IRS half of what the small business pays – as long as the business pays more than half the premium. (Details are here, page 3.)

So let’s play with the numbers, and see what it’s likely to cost a restaurateur.

For a high estimate, let’s start with my premium, $6,800 a year. In fact, let’s round it up to $7,000 to make the math cleaner. Comparable coverage would cost a 20-year-old $3,500 a year (probably less, but I’m trying to be conservative even here). The employer would have to pay for half – $1,750. And of that, the government pays for half.

So it’s not costing the employer $6,000 per year. It’s costing $875.

Run the same drill on the crappy $1,800-a-year plan, and the cost is $187.50.

Now that’s a real cost, and if you are making $50K from your business and have 50 employees, you have a real problem. Your costs will have to rise. (Of course, your competition is likely in the same position.)

But it’s more than a bit misleading to suggest that employee costs will rise 25%, as the Times data suggest. A more reasonable estimate is $875, or less than 4%. But one could make the case it will be less than 1%.

By the way, I think it’s more than a little creepy that the people handling my food don’t have health insurance.