How State Farm Lost $8B In California

Yes, they underpriced business and had poor risk management. I think we all know that.

But how did it happen? If you saw the recent Wall Street Journal article, you know that a state manager chose to prioritize making the agents happy over making State Farm money.

This led to State Farm taking on a crazy amount of growth in recent years, even though actuaries, consultants, and others were warning them the business was terribly underpriced and nobody else would touch it at those rates.

Why weren’t there safeguards and controls in place to prevent this? There probably were on paper, but they ignore the other side of risk management… managing incentives.

Principal-Agent Conflict

I’ve written about this several times before, but agency theory, the study of how to align the interest of agents (i.e. employees, specifically executives) with the interests of the corporation, is one of the most important responsibilities of a board at any company.

State Farm’s board obviously failed, as did its CEO at the time, Michael Tipsord (more on him later).

The work of thousands of employees setting rates, deciding on coverage terms, adjusting claims, etc. was ruined by one executive who chose to grow because he failed the principal-agent conflict test by apparently prioritizing his personal interest – acceptance from insurance agents – over managing the risk to the corporation.

I won’t mention his name. It’s in the article if you want to know, but we’ve all seen this plenty of times. A manager wants approval from either employees or customers or some other “stakeholder” and violates their responsibility to their employer to do no harm.

One can imagine someone working in that position at State Farm (or any large insurer) has to deal with a lot of headaches from unhappy agents if the company isn’t growing.

One can also imagine part of how one gets promoted to a role like that is by having good relationships with the agents. There is probably a lot of golf played and meals shared.

Over time, the manager comes to believe their job is to appease the agents, not the parent company halfway across the country.

It’s a bit like Stockholm Syndrome. The path of least resistance is to help the agents grow, even if it means doing irresponsible things such as growing like mad at subpar rates in wildfire zones.

I will note, the article also mentions the person who replaced him and turned off the growth and started seeking rate increases. I will mention her by name – Denise Hardin – because she deserves the accolades.

I don’t know her, but if I were at another insurer looking to fix my homeowner’s book, I’d probably give her call and see if she’s open to a new challenge.

Without her actions, State Farm likely would have lost north of $10B! It’s amazing how much difference one person can make.

Consequences

Before addressing how we prevent things like this from happening, I want to address a key part of principal-agent conflict – lack of consequences for destructive actions.

I don’t know if the former California manager is still with the company or not. It’s possible he lost his job and many would say that’s a severe consequence.

On the other hand, did he only get promoted to that role to begin with because of his ability to show growth? Did he get outsized bonuses for putting on all that bad business? So there likely were some rewards for making the wrong choices along the way.

You can imagine how one might logically conclude doing the wrong thing was the right choice. If you get rewarded in the short term for growth (higher pay, praise from agents, etc.), you may conclude it is the better strategy.

After all, if you only lose if there is a 1 in 100 or 1 in 50 event, that might be a good bet if you are 10 or 20 years from retirement.

The odds of a 1 in 50 happening over any 20 year period are only 1 in 3. It’s 1 in 5 over a ten year period. You can understand how someone would rationalize “it won’t happen on my watch”.

The Buck Didn’t Stop

All that said, this is where a strong risk management culture is supposed to prevent an individual from behaving this way. That most assuredly didn’t exist at State Farm.

The Journal article describes the many warnings from others internally. Where was the CEO? Surely, Tipsord was aware of how fast the company was growing in its largest state.

Presumably, someone had raised concerns about the risk profile of that growth and was overruled.

It’s easy to blame the employee, but the responsibility lies with the CEO. Tipsord’s job was to question this growth and put an end to it. He obviously failed to do so.

What were his consequences? There don’t appear to be any.

Now, he did retire last year. Maybe it wasn’t his choice? Maybe he was pushed out once the board realized the gravity of what they did in California?

I don’t know, but I’m going to presume it was a coincidence. He was 65 and his predecessor stepped down at the same age.

Regardless, even if he was pushed, he still won the game. He was making $25M a year running a private company with poor results and no accountability to shareholders!

Talk about winning the lottery. Remember, State Farm has posted some of the worst results in its history in recent years due to its underpriced auto book.

There was no impact on his pay for overseeing this mess. Presumably, there were similar principal-agent conflicts behind State Farm’s auto market share growth in recent years and he did nothing about those either.

Trader’s Option

In my old line of work, your pay could be clawed back if you had a bad year after a good one. That kept investors from following the “trader’s option” strategy.

The trader’s option was to seek high volatility investments because you’d rather get paid a lot one year and zero the next, than a medium amount each year investing more prudently.

Clawback was the way to reduce that principal-agent conflict. Clawback meant your worst case in a bad year wasn’t $0 but negative, since you had to return some of your prior pay.

The big banks implemented this practice as well following the financial crisis. But the insurance industry? Nope.

Insurance is filled with executives who take too much tail risk or grow too fast and still get most of their bonus in bad years. There is certainly never clawback.

Tipsord should give back half of what he made during his tenure. That only seems fair.

He didn’t earn his pay while he was there and he created unthinkable amounts of tail risk through his neglect. He shouldn’t be allowed to keep his gains from playing trader’s option.

Remedies

So how does the board of an insurer prevent large losses from self serving behavior gone wrong? It begins (and ends) with incentives.

The article made clear there were plenty of people who knew mistakes were being made. This isn’t like some Wall Street trading desk where one out of control trader went rogue and nobody realized it until it was too late.

State Farm’s institutions appear to have functioned properly. They identified the risks and raised red flags.

The problem was those red flags were ignored by the people who were supposed to see them and respond.

That means this is mostly a management failure, which against takes us back to the CEO and the Board.

While it’s easy to blame the CEO, the Board’s job is to make sure the CEO represents the best interests of the company. They failed.

If you want to know who they are, you can find their names here. They deserve as much heat as Tipsord.

Leading By Example

I would argue a CEO’s greatest impact on a company is setting the risk culture.

Every successful insurance CEO I know would have seen those growth numbers coming out of California and would have started raising questions.

They would have incentive pay for managers and underwriters tied to longer term results, so adverse development essentially acts as a clawback system.

They would also introduce volatility into their own comp. There is no better way to lead than by example. If the CEO’s pay gets cut in half after a bad year, the troops will notice that the general practices what he or she preaches.

Unfortunately, the personal lines insurers are well known for paying CEOs exorbitant packages irrespective of the quality of results. This needs to change.

Beyond making compensation more variable throughout the organization, there needs to be better options for someone to raise concerns internally.

Risk Officer Revamp

This is where the Chief Risk Officer should be reaching out across the organization and seeking information contradictory to what is coming from the product managers.

Too many CROs are happiest when focusing on their metrics. While measurement is an important part of the job, creating a strong risk culture is even more important.

The CRO should be meeting with line people constantly and drilling home messaging about what type of risks are acceptable and what aren’t.

They need to set the tone to remind all those state managers who want to be Stockholm Syndromed that their primary responsibility is to senior management, not the local agent.

Finally, just as a good state manager can’t be too close to the agents, a good Chief Risk Officer should not be too friendly with the CEO.

I would even argue the CRO should jointly report to the Board, not just to the CEO.

I know all my CEO friends will tell me that’s the worst idea ever, but that’s exactly why it’s needed. There needs to be checks and balances on the CEO too and the Board should hear from different voices.

New Leadership

I don’t know anything about Jon Farney, the new State Farm CEO. I hope part of why he got put in the hot seat is because he understands some of these changes I have outlined are needed.

However, color me skeptical. State Farm has a history of being run for the benefit of management, rather than the policyholder.

They have little interest in reform because that would reduce their perks. And with a weak board, no shareholders to worry about, and policyholders (the real owners!) having no voice, it is easy to understand why this is their attitude.

I hope Farney proves me wrong, but I’m not counting on it.