Sunday, March 30, 2008

Deciphering Auto Insurance Rates

Of all the bills I receive, the biannual renewal of my auto insurance ranks as one of the most expensive and mystifying. It is usually worth my effort to call my insurance agent and review the numbers with her over the phone. I am usually several hundred dollars richer after the phone conversation for reasons that make no sense to me.

My auto insurance bill is made complex by the fact that we own two vehicles and have four drivers. Two of the drivers are teenagers with my son having a license and my daughter a learner’s permit. While these circumstances make for a complex bill with lots of additional charges and discounts going into the final computation, my situation is in no way unusual. I’m certain that I am one of millions of fathers with teenagers at home learning to drive. But the banality of my situation makes me wonder even more about the kind of logic insurance companies employ when setting rates.

For example, each of the four drivers in our household must be “rated” on one of our vehicles. We own two vehicles that are vastly different in value. Our “family” car, a 2005 Dodge Grand Caravan, has a book value of about $10,000; while the 1997 Geo Prizm with over 140,000 miles that I use to commute to work, is worth almost nothing. I am “rated” on the Prizm, but the rest of the family was “rated” on the Caravan.

“Can I drive a vehicle that I am not rated on?” I asked.

“Yes, you can drive either vehicle,” the agent replied.

“What about my kids? Can they drive the prism?”

“Yes they can drive either vehicle; they’ll be covered.”

“What does the rating mean then?”

“We need to rate each driver on one of the vehicles to determine the rate for insurance.”

Talk about a circular definition. I then asked what would happen if my teenage drivers were “rated” on the prism. After a silence while she entered the scenario into the fields on her computer screen, she came back with a figure that was about $500 less.

“Why the big difference?’

“Because you don’t have any collision coverage on the prism.”

“It has no book value so it wouldn’t be worth insuring for collision.”

She agreed and said that she would change the policy to rate my children on the prism. She assured me that each of us could drive either vehicle and we would be covered. There would be no practical consequences to the policy change; only the rate computation would differ. I hope she knows what she is doing. But, I imagine my policy would come with dire warnings if being rated on only one vehicle meant that you couldn’t drive the other.

I then asked why I didn’t see any mention of the “good student” discount on the policy even though I had faxed report cards several weeks earlier. She wasn’t sure what happened to the faxes and admitted that the discount had not been applied. She applied it and informed me of another $300 reduction.

By the time I was finished with the phone call about $800 had vanished from my six-month policy. Not a bad outcome for 20 minutes of my time. But it was a strange kind of price negotiation. The agent’s demeanor was that of a disinterested third party (which she essentially is) rather than an advocate for her company. The price for the product was determined by some arcane formula programmed into a remote computer. Manipulating the price while staying within the rules of the policy was easy if you took the time to ask the right questions. After all, the computer only responds to numerical inputs with numerical outputs; it doesn’t know what the numbers mean either.

You wonder if anyone at the insurance company knows what these numbers mean. But asking the right questions is saving me over $1600 this year.

Joseph Ganem is a physicist and author of The Two Headed Quarter: How to See Through Deceptive Numbers and Save Money on Everything You Buy

Thursday, March 20, 2008

Fooling with Credit Scores

Bob Sullivan of has written an instructive series of articles on credit scores. He explains how credit scores determine the interest rates consumers are charged when borrowing money. Small changes in credit scoring can result in tens of thousands of additional dollars in finance charges a consumer could be required to pay on a mortgage.

What fascinates me is his explanation of how the reliance by the lenders on scoring invites people to “game the system.” Not only do the borrowers look for tricks to enhance their scores, but, the lenders use the scores to convince investors that the loans are not as risky as they actually are. It amazes me the great effort spent determining precise measures (FICO scores) of risk to lenders, while ignoring obvious problems with the bigger picture. It is a classic case of not being able to see the forest for the trees.

For example, while in the process of preparing a talk for a one-day university event this coming April, I made a simple plot using Census Bureau data of the ratio of the median home price to the median household income in the United States for the past 20 years (1987-2007). That ratio is flat at about 3 until the year 2001 when it suddenly takes off reaching nearly 5 by 2007.

Obviously the rise in home prices over the long run cannot outpace household income because a point in time will come when people no longer have the money to make the loan payments. I’m sure all those brokers at Bear Stearns invested a great deal of effort and expense modeling their risk with precise credit score data. But, a simple look at Census Bureau data would have shown them that their business model—betting that even in a foreclosure there would always be another buyer able to pay an even higher price—was doomed to failure.

Joseph Ganem is a physicist and author of The Two Headed Quarter: How to See Through Deceptive Numbers and Save Money on Everything You Buy

Monday, March 17, 2008

Refusing to Take No for an Answer

The “Do Not Call” registry has been a wonderful advancement in phone etiquette. I remember signing up the first week it became available and it resulted in a noticeable reduction in annoying sales calls. However, the calls did not reduce to zero because of a loophole that allows companies to call existing customers. This seems like a reasonable exception. For example, my bank should be able to call me if there is a problem or reasonable question to ask about my account. But recently my bank has been calling me everyday, both at home and at work, to sell me insurance. It appears that as long as I don’t buy the insurance, my bank will call to keep asking. Saying no to the offer does not end the daily calls.

The calls are a minor nuisance at work because I have caller ID. I don’t answer the phone when the bank’s number appears and a voice mail is never left. At home I do not have caller ID and it’s more of a problem. I have found an effective response to quickly end the call. When I pick up the phone and hear the short pause followed by a hesitant mispronunciation of my name in the question: “Is Joseph Ganem there?” I reply: “May I take a message.” The result is a rushed: “No, this is just a courtesy call from his bank,” followed by a quick hang-up. My wife has followed me on this and asks to take a message for me first before calling me to the phone. I need to teach my children next to ask to take a message before admitting that I’m home and interrupting me to take the phone call.

Someday, when I have the time, I would like to engage the salesperson in a conversation about the answer “no.” Is it possible to say “no” and have that answer stop the calls for a period of time longer than 24 hours? Of course, I’m aware that the salesperson doesn’t know and doesn’t care about the answer to that question. But it could lead to some interesting musings about the meaning of the word no. Because “no” does not mean “never,” does that allow the solicitor to repeat the same question with a given frequency hoping for a different response? If that is case, what is a reasonable repetition frequency? Every hour? Every day? Every week? Every month? Every year? What if I did say never? Would that make a difference or are yes and no the only allowed responses?

Does the computer dialing the phone have a repetition frequency programmed, or does it just keep cycling through the same list of numbers over and over again? I would like to know if anyone at the bank has thought or not thought about these issues and what the rational is for annoying their customers.

Joseph Ganem is a physicist and author of The Two Headed Quarter: How to See Through Deceptive Numbers and Save Money on Everything You Buy

Thursday, March 6, 2008

Rebate Run Around from T-Mobile

I spent some more time on the phone with T-Mobile this week, trying to collect the rebate promised on a cell phone upgrade. The last time I upgraded the cell phones for my family, I was given the new phone in the store. This time I was told I had to purchase the phones for $50 each and then file a claim for a $50 rebate on each one. My first reaction was to wonder why the company would go through all the expense of processing rebate claims and mailing checks when the phone could just be handed out in the store? Two months later, I know the answer. If the rules of the rebate process are made difficult, time consuming, and onerous, T-Mobile can get away with not issuing a check and simply pocketing the money.

My problems began because I needed four phones for my family plan and had to file four separate rebate claims. I needed to fill out four rebate forms, make four copies of the receipt, and cut out four barcodes from four separate boxes. The rebate forms asked for the “customer’s phone number” and because I paid for the phones and the account was in my name, I thought I was the “customer.” So I put my phone number on each of the rebate forms.

Big mistake. It turned out that the “customer” according to T-Mobile is not the person paying for the phones. The “customer” is the phone. T-Mobile wanted the number for the phone associated with the rebate claim. I received one $50 check and three rejection letters for the other claims. The reason given is that my phone number could only be used once.

I called to complain about the rejections. After navigating the voice menu I reached a real person who said she could straighten everything out by manually entering in all the correct information. I gave her the 9-digit tracking numbers for each letter I received, the 15-digit serial numbers for each phone, and each 10-digit phone number. She read back to me the new 9-digit tracking numbers for each of my three new claims. After reading off all 129 digits [3 times (9 + 15 + 10 + 9)] she seemed rather confused and put me on hold. A while later she returned to tell me she had entered some information incorrectly, but not to worry, her supervisor would fix everything and my checks would be sent.

Two weeks later I received three new rejections letters with the three new tracking numbers. So this week I repeated the same phone conversation and with a reading of all 129 digits to a different operator at the rebate center. She assured me that my checks would be sent. When I asked what was different this time, she told me that, unlike the previous operator, she had entered all the information correctly this time. I like her confidence, but, that’s a lot of digits to get correct.

I’m still waiting for the checks. When I upgraded the phones, T-Mobile insisted I sign a contract promising to pay $800 ($200 per line) if I discontinued the service within the next two years. The contract has no contingency and no penalty for T-Mobile failing to honor their rebate promise.

Joseph Ganem is a physicist and author of The Two Headed Quarter: How to See Through Deceptive Numbers and Save Money on Everything You Buy