One more thought on asset allocation and the LTB&H strategy.
An asset allocation model tempers the gains made in periods of roaring gains ("selling tech" in '99) with a practice of picking from the best of currently lesser performing classes (mid-caps and energy in '99, for example). It also forces me to consistently evaluate which of my current holdings could be sold for the reallocation decision, which is sometimes the most difficult decision in a period of generally rising prices.
The euphoria of surging prices in one asset class may blind us to assumptions which may now be stretched. By holding 30-35% max in tech, I'm forced to sell and invest in other asset classes precisely as prices surge. I'm now in the position of being forced to buy tech (adding JSDU, NTAP, BRCD, CREE. EMC) just as I was forced to sell (by writing expensive calls) a (too) small amount of each late in 2000. I didn't sell the QCOM then, because I felt opportunity was about to unfold.
My model has been 30% tech, 35% S&P/midcap/energy/etc, 30% fixed income since the 1980s. There's another 5% which can be distributed where appropriate. Energy's done well, and I'm lightening up there as that class bumps against its upper limit. That money is going back to JDSU, etc.
UF has often made the point of keeping your winners through LTB&H, and it's an important lesson. A process which forces you to examine and prune is a complement to that lesson.
Looking back, I can see a model of 65% tech since 1985 would have given me a much greater total dollar amount, but would have required me to make many more decisions, be a better market timer, and probably get a lot less sleep...
GSR |
