NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/misc/invest/17330

Text File | 1993-01-28 | 3.8 KB | 71 lines

Newsgroups: misc.invest
Path: sparky!uunet!microsoft!wingnut!randyn
From: [e-mail] (Randy Nevin)
Subject: funds: timing vs. buy&hold
Message-ID: <[e-mail]>
Date: 28 Jan 93 21:06:56 GMT
Organization: Microsoft Corp.
Lines: 61

i'd like to start a thread on buy&hold vs. market timing, especially with
mutual funds.

i wrote a small program to help me analyze mutual fund performance. input is
friday closing nav's for about the last 4 years, plus income dividend and
capital gain distributions. it calculates running averages (of a hypothetical
$10 investment) between 5 and 50 weeks, and looks for buy signals from 1% to
5% over the average, and sell signals from 1% to 5% below the average. it
selects the 'optimal' average size and buy/sell signal levels to maximize the
final investment value. right now i'm plotting about 20 funds, plus microsoft
stock (for fun). i assume that when not invested, you can get 3.5% in a money
market fund.

i've noticed that for some funds it definitely pays to switch in and out
during specific times, and for others it seems hardly worth the time. for
example, the prudent speculator leveraged fund. (this is one i specifically
selected because of its extreme volatility, and because it has won the
dubious distinction of finishing near the bottom of all mutual funds; i
wanted some inherently 'bad' funds to measure/contrast 'good' funds against.)
using buy&hold, your 4-year return would be -3.3%, so you would have lost
money. following the buy/sell signals, your return would have been +29%, and
kept you out of the fund about 30% of the time. to be fair, i haven't
gathered the distribution data for that fund, so those numbers are off, but
still i expect there is a big difference. ok, let's take the american
heritage fund -> 12.3% vs. 31.3%, and you're in money market about 40% of the
time. (these percentages are all annualized)

now with a fund like analytic optioned equity it seems hardly worthwhile to
worry about market timing. buy&hold gave 7.9% vs. 8.9% for timing. and there
was only one sell signal for a duration of 3 months during late 1990 (when
almost all the funds i track had a distinct dip). also, vanguard wellesley
(12.5% vs. 13.8%) which generated more sell signals (4). some of the other
funds i track would have you jumping into the money market maybe 10 times
over the last 4 years just to boost overall yield maybe 2%; i really question
whether it is worth trying to time such funds. occasionally i get a sell
signal, and then my next buy signal is significantly higher, so i end up
losing some ground by sitting in the money market.

now, contrast that to a volatile stock such as microsoft, where timing would
have given 46.5% vs. 52.87% for buy&hold. so you would have done better to
buy it and forget about it (for 7 years).

what do people think about this? is there some identifiable trend here? the
most obvious thing would be that certain categories of mutual funds
(aggressive growth -> american heritage, prudent speculator, 20th century
ultra) should be market timed to pick the investment period and avoid the
sharp downs, and others (income -> wellesley, growth&income?) should be used
for 401(K)'s, IRAs, college funds and general long-term buy&hold investing.
yet this doesn't seem to transfer over to a volatile stock (microsoft); does
the theory not transfer from mutual funds to stocks, or is it an anomaly of
my (non-representative) examples? has anyone else created a mechanism like
this; does it seem to work in practice? (i have not tried to use mine in the
market yet) for my buy/sell signals i'm using a simple percentage above/below
the running average; should i be using a more complex mechanism? (slope of
intersecting lines? weighted average? number of weeks above/below a
threshhold?)

comments? favorite funds?

--
Randy Nevin    [e-mail]    ..!uunet!microsoft!randyn