After reading _Worry Free Investing_, I took a serious look at I bonds. I didn't really like what I saw.
It seems that I bonds have a pathetic real return. After tax they have a .29% return for the 25% bracket, at least for now. The 1% fixed I bonds are just about 0% real after tax. For comparison, in the same tax bracket, money markets are at about .5% real after tax.
Bernstein suggested I bonds might have a negative real return, and for the higher tax brackets that does appear to be the case.
According to my simplistic calculations, if the inflation rate equals (fixedrate * taxbracket) / (1 - taxbracket) then return will be zero real. In inflation is more then you'll lose money in real terms. For the 25% bracket, the fixed break-even point for 3.1% inflation is just about 1%. If inflation is at 3.5% (the historical US average?) the fixed rate has to be about 1.17% to break even in real terms. Etc...
This all got me to thinking. Perhaps a better way to keep real returns would be to use 90 day t-bills? Theoretically, they won't sell unless the yield negates inflation and includes a positive return after tax. Perhaps this doesn't work in practice?
In any case, it doesn't seem like I bonds are the "better cash" I thought they might be.
Perhaps we just have to wait until the inverted yield curve corrects?