It's dangerous to call turning points in long-term trends, but I'm going to go out on a limb. To wit, we are very likely in the last stage of a series of cyclical mortgage refinancing booms, which began in the early 1980s. This suggests that the mortgage market will be very different in the coming decade. It will be driven primarily by purchase as opposed to refinance mortgages. This forecast is based on the cycle of long-term interest rates. After peaking at over 18% in 1981, long-term fixed mortgage interest rates have been on a dramatic, long, and irregular downward trend. They fell sharply 1981-1985 to the 10-11% range, and then in the 1990s they've continued their irregular downward trend until in the last year falling below 7%. Mortgages are unique financial obligations in that in most cases the borrower has an open option to prepay the obligation whenever rates fall enough to cover the costs of a refinance. Through time, the closing costs of a refinance have diminished. The long-term downward trend in mortgage rates has over the past two decades has created a persistent incentive for borrowers to refinance their mortgages. The irregular nature of the decline has concentrated the refinancing into periods when rates dip particularly low. It is very likely that the mortgage interest rates of the second half of 2002 will be the lowest we see for a long time to come. That of course means that mortgage rates will be at or above current levels for a long time, taking away a significant source of refinancing. This does not mean that we've seen the last of mortgage refinancing. Long-term interest rates will continue to fluctuate. In each cycle, some of the purchase mortgages made at the top of the interest rate cycle will be candidates for refinancing when rates next bottom. But, the absence of an underlying downward trend will mean each refinancing period will have less volume than has historically been the case. Indeed, if mortgage interest rates begin a long-term upward trend, the refinancing on the dips will be even weaker. I don't think a significant long-term upward trend in mortgages rates is necessarily in the offing. My best guess right now is that mortgage rates will spend quite some time fluctuating between 6% and 7.5%. In such a world, most mortgages granted at rates below 7% would very infrequently be candidates for refinancing. Those made closer to 7.5%, or at adjustable rates when fixed rates are at that level, might be attractive to refinance every three to five years. If this outlook for mortgage rates pans out (remember, no guarantee, just a reasoned guess) there would be significant implications for credit unions. First is ALM. Many credit unions have become accustomed to the long-term, fixed-rate mortgages they've made over the past few decades turning out to actually being fairly short-term assets. In other words, for most credit unions actual prepayments on mortgages have exceeded the assumptions built into ALM models. This might even have induced some to shorten the assumed maturities of mortgage portfolios. If my interest rate outlook is correct, prepayments on mortgage portfolios will be much slower in the future than in the past. Mortgages will revert in practice to being much longer-term assets. Many of the mortgages credit unions are putting on their books in the current refinancing period will be there for a long time. The second effect of reduced refinancing will be on the bottom line. If mortgage refinancing does indeed slow, so too will the periodic surges of fee and non-interest income associated with refinancing. Also, in periods of rising interest rates, interest margins will be squeezed a bit as lower-rate mortgages stay on the books and funds costs rise. Finally, market share in mortgage lending is going to be much more difficult to come by. Instead of being simply a financial transaction (reducing the interest rate and/or taking out additional equity) the mortgage application will be driven by a non-financial event: a home purchase. There will be a much weaker flow of mortgage applications. Credit unions wishing to increase market share will need to adjust strategies to focus more on the purchase market. Of course, I could be wrong in my rate outlook. To be quite honest, I thought mortgage rates had bottomed late last year at just above 6.5%. However, it's more likely I'm right this time. That's simply because there isn't that much more room for rates to fall more. Over the past two decades, mortgage rates have varied from 18% to 6%. If 10-year Treasuries fall to 0%, mortgage rates would still be at around 2%. In other words, there are only 400 basis points left of possible decline in mortgage rates. There is a lot more potential on the upside. If I am wrong, this outlook becomes even more likely next time. If mortgage rates do drop further, to say 5.5%, everything I've said here is that much more likely. It just gets delayed for a while.