2003 was an awful year for the Fidelity
Select system as far s the model portfolio was concerned. And for
the same reasons the Fund-Track model portfolios (Both single and
multi-fund) also didn’t perform well. Why? The market was more
volatile this year (seen in the standard deviations of funds rising)
but it wasn’t huge. Rather the real change seemed to be the “speed”
at which fund prices would rise and fall which resulted in sharper
and shorter moves in funds rather then the more prolonged trends of
the past. That “speed” wrecked havoc with a lot of trades this year,
particularly in the Fidelity Selects. I’m not sure why this increase
in the speed of fund moves, but I suspect higher volumes, through
increased short term trading by large players (hedge funds,
investment houses, etc.) might have had something to do with it.
Regardless, these new conditions made “upgrading” by the existing
strategy much more difficult. The one fund approach sticking to the
top 15 funds which proved very successful in the past, performed
very poorly in 2003.
This “one fund” approach led to usually being stuck in one fund
after most gain were made for it, and then slowly sinking in the
ranks awaiting a trade signal while other funds were doing well.
In that respect a new investment method has been developed with an
aim to spread investment $$ over a broader amount of funds always
buying the top fund on a weekly basis. Credit for this trading
methodology goes to Vic, who uses similar approach trading Fidelity
Select funds.
In a nutshell this new “Averaging In” approach utilizes 5 different
amounts to invest in 5 separate weeks, adjusting just once week.
This method netted a positive return of 20% this year when I back
tested it.
Back to top
Divide investing assets in 5 equal amounts and invest and upgrade
each amount in each succeeding week. Select funds have minimum of
$2,500 so a minimum of $12,500 would be needed. More realistically
at least $15,000 would be needed because an initial loss in a $2,500
amount would pull it below that level and prevent you from upgrading
it to a new select fund.
1st Week - Invest 1st amount (A) in top ranked fund for 1st week.
2nd Week - Invest 2nd amount (B) in top ranked fund for 2nd week
3rd Week - Invest 3rd amount (C) in top ranked fund for 3rd week
4th Week - Invest 4th amount (D) in top ranked fund for 4th week
5th Week - Invest 5th amount (E) in top ranked fund for 5th week
6th Week – Upgrade 1st amount (A) in top ranked fund for 6th week
7th Week – Upgrade 2nd Amount (B) in top ranked fund for 7th week.
Etc……..
There are 5 separate amounts for this 5 week rotation because
minimum (non-redemption hold time for Select funds is 30 days. 4
weeks is 28 days, so a 5 week hold time (35 days) is the minimum
needed to avoid this fee.
Sell Triggers:
If fund at top being upgraded to in is still at the top on
successive weeks, keep investing in it with additional amounts. Keep
in mind that these amounts are separate though even if they are
invested in the same fund. The first amount can be sold 5 weeks
after purchasing it; the 2nd amount can be sold on the following
week. The First In First Out rule is implemented for redemption
policies. Just keep track of amounts invested (# of shares)
Down Market- In the advent of a broadly down turning market,
as shown by either the bear market indicator (Profunds Bear fund) or
the neutral money market fund (Select Money Market) rising into the
top 5 funds, do not upgrade into a fund but rather sell back to cash
and hold until these indicators fall back in the ranks (below to 5).
As before, never buy into negative strength. I.e. never buy a fund
showing either negative short or average negative strength. If no
funds show positive strength simply go to cash (Select Money Market)
for that week. If all funds are showing negative strength either the
Money market or the Profund Bear fund will almost always be at or
near the top (the other signal to go to cash)
Remember to always keep the 5 amounts separate. E.g., if all amounts
are in cash (money market acct) because of a down market, when ranks
show market turning back positive again, invest each amount in
successive weeks – (1/5th on first week, 1/5th on 2nd week, etc…)
This is so you will always have available funds to invest every week
in a top fund as leadership continually changes.
Pros of New Trading Method: - Only need to check once a week.
Very few redemption fees assessed (only on stop loss sells).
Relatively simple to follow
Cons of new Trading Approach - Higher starting investment
amount, (Minimum of $15,000 to start)
Back to top
The
performance of the Weekly averaging model portfolio is documented in
the Performance section of
Fidelity Select Main Page, where YTD performance is updated
every quarter and 2003 performance can be seen.
This portfolio was not back-tested prior to 2002 because I
didn't feel that there was a need to back-test it further back then
that. Why? I suspect that it wouldn't have done as well as the
more focused single fund approach in years previous to 2002.
Why? Because the market conditions prior to 2002 favored a
more focused single fund approach. Then trends would be
much more sustained and smoother contrasting to the short lived and
sharper price swings that seem to be the norm in the recent past and
today. Dramatic increases in day-trading activity has certainly
added to the increased volatility and speed at which focused funds
move more recently as has increased interest in sector specific
short term investing by both small and large traders.
The ranks
on which this system for 2003 was traded can all be seen ion the "Past
Ranks" page
As stated above, for 2003 fund trends were much shorter and sharper
then in previous years. That posed problems for the existing “one
fund” rule set. This method is more adapted for the current type of
market as assets are spread among more funds and is quicker to trade
once the redemption term is overcome. As leadership changes more
frequently then in previous years in this method there will always
be $ to invest in the highest ranked fund, instead of waiting for
the fund being held fund to fall to a certain threshold and try to
find the “one” correct fund that will run. I believe this system is
able to adapt better to the frequent leadership changes that are
currently prevalent in the market and better able to capitalize on
the increase of speed in fund price changes.
The back test using this method was performed with a rigid and
consistent rule set for objectivity. But these rules can be "bent" a
little for possible advantages. For example in my tests for
simplicity and objectivity if a fund was held longer then 30 days
(because it still was in the top 5 when it’s time to upgrade came) I
would hold an additional 5 weeks until its turn in the rotation was
held. In reality after the 30 day limit is reached and one is still
holding a fund, instead of waiting until the next week that amount
gets traded to move it, (5 week later) that fund if was dropping in
the ranks quickly or showing negative strength one could trade it to
cash rather then wait.
Another
example: If a down market indicator shows up (as described above)
any fund being held past the 30 day limit, could be sold if showing
any negative strength (because it can be sold with no redemption
penalty). Again for simplicity and objectivity the model portfolios
held funds until either their turn in the rotation came up, or it
dropped over 5% and the stop loss rule was triggered.
The key
is to try an stay consistent and objective in whatever method you
deice to use.
Based on the results of this approach and the principle ion which it
was founded, the results could probably be improved by spreading
your assets even further by upgrading twice a week and therefore
using 9 amounts (2 amounts per week for 4 and 1/2 weeks = 31 days).
This would require a larger minimum investing amount of at least
$27,000 ( $3,000 per amount X 9 weeks = $27,000). I hope to back
test this method in the future soon.
Model Portfolio - A model portfolio based on this method will
be tracked for 2004 and posted to the site. It will be updated
every week.
Back to Fidelity How to
Use
Back to top
