All transactions are assumed to be at the closing net asset value. Unlike most mutual funds, the Selects are priced hourly during the market day. The maximum 3% load and Fidelity's small transaction fees are NOT included in the percent changes shown. All distributions and dividends are assumed to be reinvested, which has no meaningful effect on the rates and percentages shown. For comparison, the same period returns of the Vanguard Index 500 Trust (symbol VFINX), whose performance is close to the S&P 500 index, are also shown.
It should not be assumed that trades in the future following the Fidelity
Select Switching System will be profitable or will equal the performance of
the trades shown below.
Note: Buy prices marked by a * have been adjusted for a distribution by the fund between the purchase and sale dates. This distribution is assumed to be reinvested in the fund, which has no meaningful effect on the percent changes or overall rates of return shown. Sell or Index Buy Fund Sale/Exchange Buy Recent Percent 500 Date Purchased Date Exchange to: Price Price Change Change 1/5/98 Software 4/13/98 Brokerage 38.48* 46.85 21.8% 14.0% 1/12/98 Air Trans. 3/9/98 Brokerage 23.64 27.82 17.7% 12.3% 1/20/98 Retailing 3/2/98 Medical Delivery 46.44 50.31 8.3% 7.3% 1/26/98 Telecommun. 3/9/98 Brokerage 47.55 53.66 12.8% 10.2% 2/2/98 Computers 3/16/98 Air Trans. 38.71 39.18 1.2% 8.0% 2/9/98 Computers 3/16/98 Air Trans. 38.49 39.18 1.8% 7.0% 2/17/98 Air Trans. 3/30/98 Telecommun. 27.47 28.07 2.2% 7.9% 2/23/98 Brokerage 5/18/98 Energy Service 38.51* 42.44 10.2% 6.9% 3/2/98 Medical Delivery 4/13/98 Brokerage 28.45 28.70 0.9% 6.1% 3/9/98 Brokerage 5/18/98 Energy Service 39.63* 42.44 7.1% 5.4% 3/16/98 Air Trans. 5/4/98 Energy Service 28.73* 28.81 0.3% 4.1% 3/23/98 Utilities Growth 4/27/98 Computers 56.47* 53.81 -4.7% -0.7% 3/30/98 Telecommun. 5/11/98 Computers 56.44* 55.59 -1.5% 0.6% 4/6/98 Energy Service 5/26/98 Health Care 28.74 28.42 -1.1% -2.2% 4/13/98 Brokerage 5/18/98 Energy Service 45.02 42.86 -5.7% -0.2% 4/20/98 Brokerage 6/8/98 Retailing 44.46 43.57 -2.0% -0.5% 4/27/98 Computers 6/22/98 Software 42.31 42.19 -0.3% 1.8% 5/4/98 Energy Service 6/15/98 Retailing 31.84 24.05 -24.5% -3.8% 5/11/98 Computers 6/22/98 Software 44.69 42.19 -5.6% -0.1% 5/18/98 Energy Service 6/22/98 Software 30.06 25.67 -14.6% -0.1% 5/26/98 Health Care 7/13/98 Computers 117.09 126.53 8.1% 6.7% 6/1/98 Retailing 7/27/98 Developing Comm. 53.88 56.22 4.3% 5.3% 6/8/98 Retailing 7/27/98 Developing Comm. 56.00 56.22 0.4% 3.0% 6/15/98 Retailing 7/27/98 Developing Comm. 54.64 56.22 2.9% 6.6% 6/22/98 Software 8/3/98 Developing Comm. 46.18 45.28 -1.9% 1.0% 6/29/98 Software 8/3/98 Developing Comm. 48.71 45.28 -7.0% -2.2% 7/6/98 Technology 9/21/98 Energy Service 56.13 52.90 -5.8% -11.2% 7/13/98 Computers 10/19/98 Food & Ag 48.05 47.50 -1.1% -8.3% 7/20/98 Computers 10/19/98 Food & Ag 50.24 47.50 -5.5% -9.8% 7/27/98 Developing Comm. 9/14/98 Energy Service 23.07 20.39 -11.6% -10.0% 8/3/98 Developing Comm. 9/14/98 Energy Service 22.02 20.39 -7.4% -7.2% 8/10/98 Money Market** 10/19/98 Food & Ag 1.00 1.0091 0.9% -1.5% 8/17/98 Computers 10/19/98 Food & Ag 48.67 47.50 -2.4% -1.6% 8/24/98 Computers 10/19/98 Food & Ag 49.20 47.50 -3.5% -2.0% 8/31/98 Health Care 10/26/98 Electronics 110.55 123.68 11.9% 12.3% 9/8/98 Money Market** 10/19/98 Food & Ag 1.00 1.0055 0.5% 4.1% 9/14/98 Energy Service 11/9/98 Electronics 17.66 18.90 7.0% 10.0% 9/21/98 Energy Service 11/9/98 Electronics 16.91 18.90 11.8% 10.6% 9/28/98 Computers 3/1/99 Financial Serv. 49.84 68.42 37.3% 18.7% 10/5/98 Utilities Growth 11/9/98 Electronics 56.21 61.06 8.6% 14.5% 10/12/98 Food & Ag 12/7/98 Developing Comm. 46.45 50.66 9.1% 16.9% 10/19/98 Food & Ag 12/7/98 Developing Comm. 47.38 50.66 6.9% 9.6% 10/26/98 Electronics 3/1/99 Financial Serv. 34.24 47.82 39.7% 15.9% 11/2/98 Energy Service 12/7/98 Technology 19.62 14.80 -24.6% 4.8% 11/9/98 Electronics 3/1/99 Financial Serv. 37.52 47.82 27.5% 9.9% 11/16/98 Electronics 3/1/99 Financial Serv. 38.88 47.82 23.0% 9.3% 11/23/98 Electronics 3/1/99 Financial Serv. 41.84 47.82 14.3% 4.5% 11/30/98 Developing Comm. 3/8/99 Financial Serv. 25.13 34.64 37.8% 10.7% 12/7/98 Technology 3/8/99 Financial Serv. 73.47 88.99 21.1% 8.4% 12/14/98 Technology 3/8/99 Financial Serv. 69.55 88.99 28.0% 12.8% 12/21/98 Electronics 3/1/99 Financial Serv. 46.71 47.82 2.4% 3.1% 12/28/98 Developing Comm. 3/8/99 Financial Serv. 29.77 34.64 16.4% 5.0% ** The system treats the Money Market fund like any of the other Select funds. That means the fund will be held for a minimum of five weeks although there is no penalty for selling it within 30 days. Another reason for the minimum holding period--to avoid being shaken out by a short-term pullback--also does not apply to the money market fund. However, the five week minimum holding period would have been needed to avoid being caught in the 1987 crash in the hypothetical backtesting of the system. Investors may want to use their evaluation of current market conditions to decide whether or not to impose the five week minimum on the money market fund. The sell price of the Money Market fund shown is not the 1.00 price of the fund, but that price plus an estimate of the interest that would be earned assuming an annual rate of 4.75%. To illustrate the use of the table, the "track" starting on 1/5/98, the closest Monday to the start of 1998 in terms of market days is: Buy Fund Sale/Exchange Percent Index 500 Date Purchased Date Exchange to: Change Change 1/5/98 Software 4/13/98 Brokerage 21.8% 14.0% 4/13/98 Brokerage 5/18/98 Energy Service -5.7% -0.2% 5/18/98 Energy Service 6/22/98 Software -14.6% -0.1% 6/22/98 Software 8/3/98 Developing Comm. -1.9% 1.0% 8/3/98 Developing Comm. 9/14/98 Energy Service -7.4% -7.2% 9/14/98 Energy Service 11/9/98 Electronics 7.0% 10.0% 11/9/98 Electronics Still held(as of 1/4/99) 25.0% 9.0% -------------------------------------------------------------------- Total return for "1998" (1/5/98 - 1/4/99) 19.1% 27.7% accounting for maximum 2% annual management fee 17.1% The calculation for the total return for the track is: (1.218)(0.943)(0.854)(0.981)(0.926)(1.070)(1.250) - 1 expressed as a percent.
Go to Current Trade list
The trades listed above illustrate how the system works. It is typical that most of the profits come from one or two large gains each year. In 1996, the best trades were in Retailing, Chemicals, and the trades in Electronics and Computers that were still open at the end of the year. In 1997 the best trades were in Electronics purchased 4/14, although that trade did worse than the index fund, and in Energy Service purchased in April-July. For 1998, the best gainers were Software and Air Transportation purchased in January and Electronics, Developing Communications, Technology, and Computers purchased in September - December.
The table shows that some trades may do much worse than the market. Testing on historical data and my actual trading experience show that Select Switching should make up this gap and then some if one will stick with the system. It may take some time. You can use the updates of the table to see how tracks starting with these trades perform against the market. Because sector funds can be quite volatile, I usually recommend that an investment in the Selects using the illustrated methods be phased in using two or three tracks over a period of one or more months. See Implementation Issues under System Description for a more detailed discussion of this topic.
You can see how Pankin's personal and client accounts have performed since November 1993 under Performance. The client accounts are managed using the switching system, but most accounts are invested in more than one fund at a time. This tends to reduce both the returns and the volatility. Consequently, the client accounts' average returns for 1996 and 1997 were less than the 1996 and 1997 returns for the tracks shown above.
No claim is made that the system will perform in the future as it has in the past or as illustrated above. Also, there can be no assurances that the system will produce a profit in the future; it is possible that the system will produce losses.
December 21, 1998: The rally in the high technology sectors has continued, and the entire market has followed. Since late October, the top-ranked fund each week has been one of the high tech funds with the exception of Energy Service on Nov. 2. Despite some violent very short-term reversals, these funds have shown outstanding gains so far that are well above the broad market averages. Because the three tracks shown above all switched into Electronics on Nov. 9, the outstanding performance of that fund has enabled the tracks to close a good portion of the gaps between their performance and that of the index fund. In the past few years, the high technology sectors have shown long (over six months) and strong (fund gains above 40%) trends more than once. Followers of the system certainly hope that will again be the case.
In some respects 1998 has fit the pattern of trading system experience and hypothetical backtesting, but in other ways it has not. It appears that the current Electronics trade will remain open until some time in 1999. Assuming it finishes with a good gain, the track starting on 1/5/98 will have had two large winners--the first and last trades, which is typical. Also, the seven trades for the year is at the high end of the expected range of 5 to 7. However, the four consecutive losing trades is unusual, and one of them was much larger than the typical losing trade. All of that nets out to subpar returns (compared to the index fund) for the year-to-date. The alternative 1998 track, which started on 12/29/97 and which can be seen in the two longer tracks shown above, has been quite a bit different from the historical norms. There have been eight trades, only one of which was a large gainer. Although there have been only two losing trades, they were quite large, which has resulted in the track showing a small loss for the year. This shows again that the choice of starting week can make a substantial difference in initial returns. Eventually, tracks will merge, as these did on 9/14, and the results will be the same from then on.
November 9, 1998: Energy Service purchased on November 2 is held because it has not yet met the minimum five week holding period. It is not a good sign to get a sell signal for a fund, like Energy Service, that was purchased one week ago after having been purchased several weeks earlier. The basic system calls for holding all purchases for a minimum of five weeks. However, given the recent behavior of Energy Service, which is a very volatile fund, it may be a good idea to use a "stop-loss" on the November 2 purchase. My analysis shows that the stop level that is best for all Select funds is down 7-8% from the purchase price. Since Energy Service can be extremely volatile, aggressive and risk tolerant investors may want to give it a little more room.
The Energy Service trade just closed out in the tracks shown above, which was purchased on September 14, shows some of the difficulties dealing with volatile markets. In that fund's gyrations, at one point its price had fallen below 14, which would certainly have caused any stops to have been hit well above that price. Then the fund rallied strongly as the price approached 20. The listing above shows a decent profit on the trade, but it is unlikely that any buyer of the fund would have been pleased by its action, and few would have stayed aboard for the whole ride.
The market appears to have made a bottom in October. Since November - January is historically the best three month period for stocks and since the presidential cycle is best in the third year after the election (1999), there is a good chance that the rally will continue and that we seen the lows for the decline that started in July. Naturally, this opinion is worth just about what you paid to get it since I claim no ability as a short-term market forecaster.
The high technology sectors now seem to be leading the market rally. Electronics is up almost 10% in the past two weeks, and other high tech Select funds such as Computers and Software are also doing well. These sectors often lead major market moves up, so this is another good sign.
October 19, 1998: Computers purchased on September 28 is held because it has not yet met the minimum five week holding period.
The increased volatility combined with large and fast moves both up and down has played havoc with the nerves of those who follow the Select Switching System. There seem to be two reasonable alternatives. One is to stop trading sector funds until the market calms down and appears to be in a definite uptrend. Using a suitable timing model is a way to accomplish this. However, timing models are far from perfect, so using one may well add another type of frustration. The second alternative is to continue using the system, probably incorporating stop-loss tactics. By doing so, one is likely to participate fully in the next sustained market uptrend, but at the risk of continued up and down volatility. For example, Energy Service has moved between 18 and 13.50 and nearly back more than once in only a few days per move. On October 15-16, the fund gained over 17%!; but that recovered an earlier decline and the fund gave up some ground since then. It is definitely not for the faint of heart (or stomach).
Some technical indicators are showing that the worst is over and the long bull market is about to resume. I don't go in for short-term market forecasts, so I don't act based on those indicators. One such indicator is generated by the system. The Money Market fund is a sell for this week, which indicates that the other Select funds are starting to show greater strength. The large table above shows that buying the Money Market fund when indicated would not necessarily have improved one's investment returns. However, given the market ups and downs, it likely would have improved one's ability to "sleep at night." The punch line to an old joke--"sell down to the sleeping point"--is still good advice for those who investments worry them to a great extent.
September 2, 1998: The tables above do not reflect the use of stop-loss tactics. This is discussed in the June 15 commentary below. Then I said I would not incorporate stops into the tables in order to add even more complications, but I am now reconsidering that decision. I will probably add yet another table (gasp!) showing how stops might have worked once I figure out how to do it in a way that is not too complex and presents a fair picture. Fairness is an issue since it is easy to pick stop levels after the fact that make the past results look very good.
In practice, I had sold all Select equity funds held in personal and client accounts by the close on Friday August 28. The proceeds were placed in a money market fund. I did not avoid all of the drop late last week, but did avoid some of it and all of Monday's plunge. It should be noted that this is only half of the story because at some point I will buy back into the market. If that is at levels that are meaningfully higher than the levels at which I sold, then a nasty drawdown will have been avoided, but overall profits will be less.
My research into stops shows that a 7% intial stop and a 10% trailing stop (sell if the fund drops that much from a high) are reasonable levels. This research is not based on simulated trading, which could be quite sensitive to a few instances when stops apply. Instead, my analysis is based on the behavior of every potential trade generated by the system, except for money market trades, from January 1987 through June 1998. Applying these levels to the track starting January 2, 1996 with a reasonable set of trading rules for handling cases where the stops are triggered yields the following results:
Selects Index System Fund ------- ----- Total return as of 8/31/98 118.0% 76.5%* Annualized 34.6% 24.2% accounting for maximum 2% annual management fee 32.6% Return for 1/2/96 - 12/31/96 of the track 36.6% 21.9% accounting for maximum 2% annual management fee 34.6% Return for 12/31/96 - 12/31/97 of the track 45.5% 33.2% accounting for maximum 2% annual management fee 43.5% Return for 12/31/97 - 8/31/98 of the track 9.7% 5.4%* accounting for maximum 2% annual management fee 8.4% * The Selects System returns assume a sale at 10:00AM on 8/31/98 when the market was down only slightly. It is fairer to use the 8/28 value of the index fund rather than the much lower 8/31 value, which is the case in this table. The above values are partially hypothetical since stop tactics were not in use for the entire period. You should not assume that future returns will be similar or that they will be profitable.
These values show that the use of stop-loss tactics can make a significant difference. In some cases, which I have found in backtesting the system from the start of 1987, stops can hurt quite a bit. They are not a panacea, but at the levels suggested, they should avoid the worst drops without hurting long-term average returns. As the comparison for the last two plus years shows, stops can improve returns in certain circumstances.
August 3, 1998: For most of July it looked like sector funds were returning to their historical tendency to have trends that can be traded profitably. However, late in the month, as the whole market started to slide, the technology oriented funds that had been the strongest fell sharply wiping some healthy profits on the current trades. This behavior was typical of what we have seen since last November. Consequently, the Select Switching Strategy continues to trail the broad market by a large amount for 1998, and the poor performance so far this year has brought the longer term (since 1996) track well below the index fund.
Note that for 1996 and 1997 the sector rotation strategy did better than the index fund, which is typical of backtested results since 1987 and actual trading since late 1993. The important question is whether the longer term historical trending behavior of sector funds will resume in the not too distant future or if the recent action is what we are going to see for the most part in the future. I don't think we can answer that question until more time passes. Certainly, eight months is not enough to conclude that sector rotation has stopped working. Unfortunately, sticking with it is becoming increasingly painful. If one follows my recommendation that no more than 20-25% of the portfolio be in sector funds, then the pain is lessened by having investments in other areas that should be doing better than sector funds so far this year.
June 15, 1998: Things have gone from bad to much worse, which requires some explanation and a change in tactics. Surprisingly, the system did not signal a switch out of the Energy Service purchased on May 4 (the purchase on May 18 had and has not yet been held for the five week minimum holding period) on June 8. That failure to switch proved to be quite costly because the fund dropped by an astounding 13.1% in the week from June 8 to June 15. With the exception of the 1987 crash period, this size drop in a Select fund is unprecendented and is even worse than what happened to the same fund last November. The overall market was weak, but not exceptionally so. For example, on June 15 when the broad market indices fell a bit over 2%, Energy Service fell over 6%.
Funds in the technology sectors have also been hit hard recently as can be seen in the losses shown for Computers shown in the table above (-10.2% for the May 11 purchase as of June 15). Again surprisingly, the system calls for holding Computers for at least another week although it has been held for the five week minimum period. In the past, the system has usually been able to get out of weak funds in weak markets, but the extremely choppy (sharp moves both up and down) action has apparently confounded it.
It is apparent that volatility of the market as a whole and sector funds in particular has been increasing since the end of 1996, and this is shown by statistical measures such as the standard deviation of the average weekly changes of the Select funds. Since markets and individual stocks and mutual funds tend to fall faster than they rise, increased volatility usually affects losses more than gains. Prior to the recent Energy Service debacle (-24.5% on the 5/4/98 purchase), no trade generated by the system had lost more than about 15% with the exception of some weeks prior to the 1987 crash. Moreover, such potential losses were relatively infrequent. However, in 1997-98, the frequency of larger negative results has been increasing.
Given the increased volatility and the consequent increased risk, I believe a change in tactics is appropriate. In particular, I now think that a stop-loss should be used on all trades. That means setting a level below the initial purchase price at which the fund will be sold regardless of what the system says or whether there will be a short-term, 0.75% penalty on the sale. Given the mechanics, that means making the calculation after the close and placing the sell order for 10:00 AM execution the next market day. The fund price will probably be different, perhaps lower, when the sale is made. Based on my research, the appropriate stop-loss level is down 6-8% from the entry price. With the process described above and the likely penalty, the actual loss will probably be a percent or two greater.
Previously, my thinking was that the purpose of stop-losses at this level was to avoid situations like the 1987 crash. Otherwise, they would be relatively infrequent, and their long-term effect on investment returns would probably be small. However, I now think that stops may be become more frequent. There is not yet enough evidence whether they will actually improve returns, but avoiding losses like the ones in Energy Service is worthwhile from a psychological standpoint. In this instance, the stop-loss sale, including the short-term penalty, would have been about 10% during the week on May 18 (when the system was signaling that Energy Service was a buy before it started its fast plunge). Presumably one moved to the money market fund and then re-entered the next week by buying Health Care because the market as a whole was not particularly weak. That fund is down less than a percent as of June 15, so in this case, the stop would have saved about 14% of the 24.5% loss. Future instances may not work as well, but it only takes a couple of losses like the current one to "beat one down" and cause the abandonment of what I still think is a good trading system despite the current poor performance.
The results shown on this page will not include stop-losses. Although I recommend that you use them, to include them would add another level of complication to the far-from-simple tables here.
Another tactic, which I am currently studying, is the use of a trailing or retracement stop. This is similar to a stop-loss except that the drop is measured not from the purchase price, but from the highest price after the purchase was made. If the fund has gone up enough, such a tactic can preserve some of the profit. The downside is that sector funds that have risen can pull back by more than a small amount before resuming their sometimes spectacular climbs. This happened with Electronics more than once in late 1996. A trailing stop would have caused a sale and a switch into a fund that did not perform as well. However, trailing stops may reduce drawdowns over the longer term. One certainly would have helped last November with Energy Service. One study by someone else suggests that 8-10% is the appropriate range for a trailing stop.
To end this lengthly commentary on a more upbeat note, with the possible exception of the first two weeks of the year, we have not yet seen any large gainers this year. The history of the system has been that any track has one or two big gainers each year, which produce most of the profits. If 1998 fits the pattern, the system performance will look a lot better six months from now. Of course, the standard disclaimer that there are no assurances that this will happen and that things can get worse always applies.
May 26, 1998: The system continues to struggle. Funds that show enough strength to rise to the top of the rankings have been turning around and falling rapidly soon thereafter. This has been particularly true for Energy Service, which is one of the more volatile Select funds. In the past, its trends have tended to be longer lived, which has resulted in some spectacular gains (see 1997 for an example), but with a considerable amount of risk. Clearly, sector fund trading is not for the faint of heart. It is times like the present that test the discipline required to execute the system.
It may be comforting to know that the historical backtesting that resulted in the development of the Selects Switching System contained a period that is similar in several aspects to the current times. The system was out of the market for the October 1987 crash and bought back in the following December. From then through August 1988, it did not perform well while the market was rising. For the entire year 1988, the system hypothetically would have lost about 9%, even after recovering somewhat in the last four months of the year. The index fund gained about 16% in 1988. Since 1988, the Select Switching System performed quite well (hypothetically until I started trading it in November 1993) until last November.
Obviously, there is no guarantee that the system will resume its market beating performance, but the 1988 behavior tells me there can be extended periods when the system works poorly. If we are investing with a longer term perspective (at least five years and preferably ten or more), then such periods become less troublesome when viewed in this context. It is possible that the system will not resume its historical performance any time in the near future. We have to live with that possibility. Unfortunately, it takes far longer than we would like to determine if an investment method is no longer valid or desireable. That is another reason for a long-term investment horizon. Those with relatively short horizons (less than five years until they need the money invested) should always examine carefully the levels of risk they can assume, which are determined by how much money they will need their investments to provide. If that amount is close to the current value of the investments, then the best course of action is likely to be very little or no investment in stocks and equity mutual funds.
April 20, 1998: Despite the great market so far this year, it is been somewhat of a mixed bag for the Select Switching System. The track starting 1/5/98, which is illustrated above, is beating the index fund due to the strong performance of Software. However, the longer term tracks are trailing the market by a fair amount so far this year. Such behavior is not unusual. A fund purchased or a track started one week may do a lot better or a lot worse than one purchased or started a week earlier or a week later.
Is this a problem or a weakness in the system? Not really. While it would be great if there were a way to tell the best starting weeks in advance, I don't know of any such method. After some time, usually less than a year, tracks merge if the system is followed explicitly. That means, that the 1999 performance, for example, will probably be the same no matter when in early 1998 one started a track. Since I advocate long range thinking, one should not be too concerned whether the first few trades do better or worse than their neighbors in time.
One method to smooth out the initial performance, which is discussed on the implementation issues page, is to phase in the start. Divide the funds to be invested into two to four pools and use each pool to start a new track at a regular intervals, say a month apart. Let the tracks merge as called for by the system. That method has the advantage of not jumping in with all your funds at a market top. Of course, if the market keeps going up, some of your money will be on the sidelines.
March 9, 1998: The past few days have been turbulent ones for the high technology sectors. In particular, Computers, which is the current fund in both of the tracks shown above, fell by 6.5% in the past week. That drop erased the advantages the tracks had been showing in comparison to the index fund. The fall in the high tech areas was set off by Intel's and Compaq's announcements that first quarter earnings and revenues would be well below expectations. This type of volatility is typical for investments in sector funds, and one has to be able to live with it. In the past, these types of shocks, particularly in the high techs, have often proved to be short lived, and the funds have quickly recovered and gone to higher levels. We should know relatively soon whether or not such will happen this time.
Note that Air Transportation purchased on 2/17/98 should not be sold since it has not been held for the five week minimum. Since both of them have done well and outperformed the strong index fund, it is curious that the January purchase of Air Transportation and Telecommunications are to be sold . Even more curious is that Computers purchased on 2/2/98, which has been held for the five week minimum and which is down in comparison to the index fund, is to be held for another week. There will be times when the system seems to do "silly" things, and this may well be one of them. At such times it is tempting to use one's "judgment" and override the system. I do not advocate such behavior. While overriding the system may prove to be correct this and other times, the problem is that it won't be long until the overrides are fairly frequent and, in effect, the system has been abandoned. For most investors, this is a prescription for poor results. While all systems are far from perfect, in the long run they do better than most investors using judgment to make ad hoc decisions. As I have said before, the key is sticking with the system, which is especially important when the system does not seem to be working well.
February 9, 1998: The merger of two tracks is possible this week since both Multimedia and Utilities Growth are due to be sold and switched into Computers. The top four funds for this week are all in the high technology sector, so buying the number two ranked fund will not really maintain diversification. In such a case, if one had owned the two funds to be switched and wanted to maintain diversification, about the only alternative is placing the proceeds from one of the sales into a money market fund. When a non-high tech fund became top ranked (or possibly ranked number two), then the money market could be switched into that fund.
The Multimedia trade, which is part of the tracks illustrated above turned in a small profit and did not do as well as the index fund. Such behavior is typical of the system. In an "average" year, there will be about six completed trades. One or two should be substantial gainers that outperform the market by a large margin, one or two will lose (a small amount we hope), and the rest should produce fair gains that in many cases will lag the market. The past two years fit this pattern for the most part. The number of trades has been a bit lower due to long trades in Electronics and Energy Service. Some of the losing trades have been larger than we would have liked, but overall the system has done well and produced market beating returns. Of course, nothing is guaranteed and the system may lose money in the future.
Go to Implementation Issues