Lu:
1. This is sometimes obvious, usually not. MSFT, INTC have this
quality, IMHO. ORCL had this quality 5 years ago (future
beyond 1-2 years unknown as MSFT moves into databases). I'd
recommend you read some books.
2. Growth projections I get from my broker, and by looking at
history, and (where available) by looking at independent
market analysis (e.g. Gartner Group, etc.) Some 5-year
compound annual growth rates on my "fab five":
MSFT: 20-25% (broker), 25-30% (me)
INTC: 20%
SUNW: 20%
CSCO: 30-50%
ORCL: 30-40%
3. Buying is easy. If your target is well-chosen, you won't go
wrong long-term. I use a valuation system based on projecting
the growth rate out 5 years (optionally 10 if you really like
the stock) and discounting it back versus a market average.
The following is *not* investment advice, but my personal
opinion on some stocks I follow:
MSFT: accumulate (fully priced at this level, buy fixed dollar
amounts over many months to enter).
INTC: accumulate (as per MSFT).
SUNW: hold (price is fair, but franchise may be threatened
beyond 2 years).
CSCO: buy (price is fair, growth rate strong).
ORCL: hold (fully priced).
NSCP: sell (very overvalued).
I watch valuations on my favourites (such as MSFT) primarly so
that I can buy aggressively when they're cheap (bought MSFT big
at $84 in January based on this).
For what it's worth, my personal valuation table based on
current interest rate/inflation environment:
10% growth: 12.0 P/E (forever)
15% growth: 14.9 P/E (5yr), 18.7 P/E (10yr)
20% growth: 18.5 P/E (5yr), 28.6 P/E (10yr)
25% growth: 22.7 P/E (5yr), 43.0 P/E (10yr)
30% growth: 27.7 P/E (5yr)
35% growth: 33.4 P/E (5yr)
40% growth: 40.1 P/E (5yr)
45% growth: 47.8 P/E (5yr)
50% growth: 56.6 P/E (5yr)
Note that any growth rate greater than the market average can
be made to justify an arbitrarily high P/E today, if you're
willing to go out far enough. For instance, if you believe
MSFT can sustain 25% growth for 9 years, (or 20% for 13
years) you can justify today's 38 P/E. For growth rates
above 25%, I'm unwilling to look beyond 5 years. Above 50%
growth, I make up a custom projection out 5-10 years that
gradually soft-lands the growth rate to something sustainable,
then discount it back based on a soft-landing of the P/E.
4. Usually, I don't sell unless a stock becomes incredibly
overvalued. For instance, I'd be selling KO (Coca-Cola) at
these prices (for a 40 P/E, I'd rather take the proceeds
of selling KO and buy MSFT). But, if I'd owned HWP before
it fell, I would not have sold even though it was overvalued
at 25 P/E (vs 15% growth). "Incredibly overvalued" means
the price is 50-100% over fair, meaning I can exit, pay my
taxes, and still be ahead after reinvesting. This rarely
happens (and never happens in the case of MSFT).
Sometimes, I sell because I need to rebalance my portfolio.
For example, CSCO's growth rate looks like it will slow to
below 80-90%, so I need to find something else to occupy
the "high-growth speculative" portion of my portfolio.
Occasionally, I will sell if I feel that the "dominant market
position" or some other intangible quality is in doubt. This
is very, very difficult to know. Often, these doubts are wrong.
For instance, I have vague unease about the following:
1) ORCL's dominance of databases as MSFT moves into this area
2) SUNW's dominance of servers as Wintel moves into this area
3) CSCO's high margins as competitors move into networking
There is no evidence yet of any of these problems. Frankly,
if you're already in a stock, it's usually better to wait
until there is solid evidence, let the stock take a hit, and get
out, rather than trying to anticipate the trend. (Try to
anticipate, and you'll likely be either totally wrong, or years
too early -- either way, you'll miss a lot of upside on the
stock.) If I must anticipate, I try taking 1/4, 1/3, or 1/2
my profits "off the table", as I did with ORCL a few months
ago. (Note, however, that these sorts of doubts are very
relevant to making a *buying* decision.)
Good Luck! |
