While direct kickbacks are illegal under the Real Estate Settlement and Procedures Act (RESPA), there are numerous ways to legitimize them. One that has become common among large lenders is to establish a reinsurance affiliate that shares the premiums on insurance sold to the lender's customers. This is kosher under RESPA, since the affiliate also shares the risk. The reality, however, is that reinsurance deals are disguised (and costly) kickbacks.

Traditional mortgage insurance has another unholy feature – the insurance runs on well past the time that it is really needed. Since the insurance protects the lender but the borrower pays for it, the lender has no incentive to terminate the policy when the risk becomes minimal. In 1999, Congress finally decided to do something about this, establishing mandatory termination rules. The rules, however, are extremely complex and difficult for borrowers to navigate.

Fortunately, there are a lot of lenders in our system, and some of them have little stake in the traditional mortgage insurance system. When they discovered a few years ago that they could obtain a competitive advantage by offering combination first and second mortgages instead, they jumped at it. For example, to the borrower who could only put 5 percent down, they offered an 80 percent first mortgage plus a 15 percent second, in lieu of a 95 percent first mortgage with mortgage insurance.

These came to be called "piggybacks." While the second mortgage has a higher rate than the first, the higher rate is paid only on the second mortgage and the interest is deductible. Premiums on traditional mortgage insurance are paid on the entire first mortgage, and are not deductible.

Within just a few years, piggybacks became established as a major alternative to traditional mortgage insurance. With their traditional business shrinking at an alarming rate, the insurers have been under enormous pressure to develop counter-measures. Lender-pay insurance is it.

Under lender-pay insurance, the lender pays the premium and charges the borrower for it in the rate. This is better than traditional mortgage insurance because lenders have an incentive to pay as little as possible for the insurance, rather than to benefit as much as possible from their referral power. Since lenders must compete in terms of interest rate, the borrower ultimately will get the benefit of lower insurance premiums.

The rate increment lasts as long as the mortgage, but that is also true of the second-mortgage part of piggyback arrangements. In addition, lender-pay insurance is simpler: one loan, one rate. Piggybacks usually involve an incremental upfront fee, and the second mortgage can be a different type of instrument than the first mortgage. Frequently, it is an adjustable-rate mortgage of some type, which makes the package more difficult for borrowers to assess.

In my view, a system based on lender-pay insurance will work better than one using traditional insurance or piggybacks. This does not imply, however, that in our existing system that offers all three choices, borrowers will always do better with lender-pay insurance.

Most loan providers charge what the market will bear, which means that you can easily overpay for any of the options. Contrary to what you may hear from a loan provider, there is no general answer to the question of which approach is less costly to the borrower. There are only specific answers to individual deals, and the answer can vary from deal to deal.

To help with this problem, I developed two calculators, which are on my Web site. Calculator 14a compares the costs of traditional (borrower-pay) insurance and lender-pay insurance. Calculator 13a compares the costs of a piggyback deal and lender-pay insurance. Calculators are an excellent defense against high-powered sales pitches.