Your logic in comparing the index plus margin, called the “fully-indexed rate”, is sound. But it is subject to a caveat that I will explain below.
On an adjustable rate mortgage (ARM ) with an initial interest rate that holds only for a short period, the fully-indexed rate is more important than the initial rate. The fully-indexed rate indicates what will happen to the rate at the end of the initial rate period, which in your case is one year.
On May 24, when I checked the rates, the most recent value of the COSI was 5.00% while the one-year Treasury was at 6.29%. Adding the margins of 3% and 2%, the fully-indexed rate was 8% for the COSI and 8.29% for the one-year Treasury.
If you took a 7% ARM on May 24 and the rate indexes did not change during the following year, the rate on the COSI ARM would rise to 8%, and the rate on the one-year Treasury would rise to 8.29%. This is your reason for preferring the COSI.
Unfortunately, it is not that simple.
If the Treasury index stays at 6.29% over the next year, the COSI index would rise. That’s because as of May 24, the COSI index did not fully reflect the increase in market rates that has occurred since the beginning of this year. The rise in COSI could easily eliminate the .29% difference in fully adjusted rates by the end of the year.
The COSI is a monthly index that lags the market. Part of this is simply a reporting lag. On May 24, the most recent COSI value was 5.00%, but it was for March. Treasury indexes are available on a daily basis, and the 6.29% value available on May 24 was for May 22.
COSI also is subject to a market adjustment lag. It is an average of the interest rates paid by west coast savings institutions on all their fund sources– transaction accounts, savings deposits, certificates of deposit, and borrowings of various types. When market rates change, the rates on these fund sources adjust slowly. The rates on certificates of deposit, for example, change only as new certificates are written at higher rates while the older ones carrying lower rates mature.
The Treasury indexes, in contrast, always reflect the current market because they are affected by what investors are willing to pay at the moment when they buy or sell. Some observers would say they are the market.
Some investment bankers have developed COSI forecasting models for investors. The models estimate how past and projected future changes in market rates will affect future COSI values. This type of information is not available to consumers.
Because of the reporting and market adjustment lags, the COSI index is much less volatile then Treasury indexes. This is a reason for borrowers to prefer it, assuming other features of the ARM are the same. But the lags that stabilize COSI also make it difficult to compare fully indexed rates of COSI ARMs with ARMs tied to Treasury and similar indexes.
For this reason, borrowers should compare fully indexed rates only among ARMs using the same or similar type of index. You can compare one COSI ARM with another, or with a similar COSI (cost of savings index) ARM. Treasury and LIBOR indexes are also comparable. 12 MTA, which is a 12-month average of one-year Treasury securities, is a hybrid measure that is most comparable to COSI.