7/29/2019 aiq

1/8

AIQ Opening BellM o n t h l yVOL . 5 ISSUE 8 AUGUST 1996IN THIS ISSUE

DR. J. D. SMITH

THE TREND SCORE IS YOURFRIENDBy Dr. J. D. Smith

F e a t u r e

AIQ's Trend Score -The Trend Direction & Strength

Indicator ............................... 1

S e c t i o n s

TradingExpert's New MarketBreadth Builder ................... 6

Data Maintenance ............... 7

Market Review ..................... 8

The Opening Bell Monthly

is a publication of

AIQ Incorporated

David Vomund, Chief Analyst

P.O. Box 7530

Incline Village, Nevada 89452

Putting it All Together continued on page

PUTTING IT ALL TOGETHER

Many of our most popularclichs involve the trend inprices, be it for the market,

sectors, stocks, or indexes. Youve heardthese clichs The trend is yourfriend, Trade the trend, Dont go

against the trend. Some very good rulescan be gleaned from these concepts forour trading process rules that havestood the test oftime andscrutiny.

In order toimplementtrading rulesinvolving thetrend, we needto be able to

measure thetrend. We mustknow thedirection andthe strength of the trend. For this reason,we developed an indicator called theTrend Score which measures trenddirection and trend strength. Trend Scoreis found on AIQ stock, group, and sectorreports.

The Trend Score has a range of valuesfrom +100 to -100, where +100 is a strong

Trend Score is exactly whatthe Expert Rating is not.

The Expert Rating iscountertrend. Trend Score istrend following. The Expert

Rating marks the exceptional

events . . . Trend Score measuresnormal indicator behavior.

uptrend and -100is a strongdowntrend.Values aroundzero indicate that no trend is present.

Before we get into the details of theTrend Score, lets look at an example of

how rules determine AIQ Expert RatingOn June 7 of this year, the market timing

expert systemgenerated adownsideExpert Ratingfor the market o3 - 97, as shownin Figure 1(page 2). As faas the Dow waconcerned, the

signal was earlby about threeweeks. But thebroader market

started to weaken at that point in timeand continued to correct up to the timethis article was written.

Figure 2 shows the three rules thatfired on June 7 to cause the signal. Thefirst rule revealed that the smoothedAdvance/Decline Line, which representmarket breadth, had turned negative,

7/29/2019 aiq

2/82 AUGUST 1996

AIQ Opening Bell

PUTTING IT ALL TOGETHER continued . . .

Opening Bell MonthlyG.R. Barbor, EditorP.O. Box 7530Incline Village, NV 89452

AIQ Opening Bell Monthly does not

intend to make trading recommendations,nor do we publish, keep or claim any track

records. It is designed as a serious tool to

aid investors in their trading decisions

through the use of AIQ software and an

increased familiarity with technical

indicators and trading strategies. AIQ

reserves the right to use or edit submis-

sions.

For subscription information, phone1-800-332-2999 or 1-702-831-2999.

1993-1996, AIQ Systems

PLEASE SEND CORRESPONDENCE TO:

that down volume was greater than upvolume, and average declining issueswere greater than the average advanc-ing issues. Together these indicationscreated a bearish sign.

The next rule fired because thePrice Phase indicated advancing prices(the Dow advanced 30 points that day)but Volume Accumulation Percentageindicated that money was flowing fromthe market. This nonconformationprovided another bearish sign.

The last rule fired because theHigh/Low indicator, a very stable andconsistent indicator measuring marketbreadth, reversed direction and startedmoving in a bearish direction. Anotherbearish sign.

Each of the rules in the expertsystem has an assigned weight, andwhen a rule fires the rule weight iscombined with other weights fromother rules into an Expert Rating. On

June 7, three bearish rules fired and theresult was the 3 - 97 bearish ExpertRating for the market. Later, I willcompare this type of expert systemwith the special type of expert systemthat is used in determining the TrendScore.

Returning to our example, wewould need some confirmation beforeturning completely bearish at this pointin time. The Weighted Action List onthis date had an Up/Down Signal

Figure 2

Ratio of 22 - 78 bearish certainly, butnot enough to confirm the ExpertRating.

Using Trend Score to Confirm aMarket Expert Rating

The Daily Group Analysis reportfor the AIQ Pyramid group structurefor June 7 is shown in Figure 3. AIQsgroup and sector reports are sorted byTrend Score. The Score at the top ofthe June 7 group report is 68 - 32,

meaning that 68% of the groups in theAIQ Pyramid are in an uptrend. TheTrend Score was computed for eachgroup and 68% of the groups were inan uptrend while 32% were headingdown. Thats still bullish.

The Delta value at the top of thereport is 28 - 72, indicating that 72% ofthe groups had decreasing TrendScores from the previous period. Nowthat is bearish. Delta, a very goodbroad market indicator, is indicating

Figure 1

7/29/2019 aiq

3/8AUGUST 1996

AIQ Opening Be

PUTTING IT ALL TOGETHER continued . . .

Figure 3that a downward price movement is already inprogress.

A nimble trader could look at the groups atthe bottom of Figure 3 for stocks to short. A moreconservative trader would at the very least tightenstops, and wait for further confirmation of theExpert Rating.

A week later, the Dow had dropped 50 points.The indicator barometer on the market screen wasshowing red indicators and 51% of the groups inthe AIQ Pyramid were now in a downtrend.Figure 4 is the bottom section of the GroupAnalysis report for June 14. The correction wasfirmly in place. The nimble trader was countingprofits, and the more conservative trader wasstarting to clear long positions as stops weretriggered and looking for short positions or atleast for a money market account to weather the

coming storm.The Group Analysis report in Figure 4 is the

perfect place to look for stocks to short. Thegroups that were in strong downtrends a weekprior are still there but additional groups haverotated downward and offer opportunities forshort positions.

As more and more groups show strongdowntrends, the DTS (Delta Trend Score) columnbecomes more important. DTS shows the changein the Trend Score from the previous period. Weuse this to show which groups are new arrivals to

the extreme ends of the report. For example,Railroads in Figure 4 has a Trend Score of -100, astrong downtrend, with a Delta of -26, whichshows that this group just made the bottom of thereport on June 14. The group has rotated downand is currently in the early stages of a strongdowntrend.

Figure 5 shows another of the groups from theJune 14 Group Analysis report. Precious Metalswas in a strong downtrend the prior week,according to its Trend Score, and still is. As Figure5 shows, opportunities in the Precious MetalsGroup are still available in mid-June.

This example shows how Trend Score can beused to confirm a market timing Expert Rating,and then used to select groups offering the mostopportunities for profits in the prevailing marketdirection.

How Trend Score Is Determined

Now, exactly what is the Trend Score. TrendScore is exactly what the Expert Rating is not. TheExpert Rating is countertrend. Trend Score is

Putting It All Together continued on page 4

Figure 4

Figure 5

7/29/2019 aiq

4/84 AUGUST 1996

AIQ Opening Bell

PUTTING IT ALL TOGETHER continued . . .

trend following. The ExpertRating marks the exceptionalevents among the indicators.Trend Score measures normalindicator behavior. And theExpert Rating is periodic, while

Trend Score is a continuousmeasurement.

Trend Score is based on themovement of prices, on the direction of the trend in prices and thestrength of the trend, on money flow, and on supply and demand.Prices are measured by the Price Phase indicator. Trend is measuredby the Directional Movement Index and ADX. Money flow is mea-sured by Volume Accumulation Percentage. And supply and demandis measured by the Positive Volume indicator.

Each of these indicators is placed in a rule and the weights fromeach of these rules is combined into the TrendScore. Here is an example of an Expert Rating

rule and a Trend Score rule.Rule 1 (Figure 6) is an Expert Rating rule. It

says that if closing prices are at a 21-day highand if Volume Accumulation Percentage is lessthan zero, then the rule fires down. The weightassigned to the Expert Rating is the reliability ofthis rule to predict downward movements inprices. This is a countertrend rule. Prices arehigh but there is distribution, or money isflowing from the stock or index or market.

This is a periodic rule that will fire onlyonce in a great while. If prices are at new highs

and accumulation is occurring, it is normalbehavior and the rule shown in Figure 6 doesnot fire. This rule fires only on the exceptionalevent when prices are at new highs and distri-bution is taking place.

Compare and contrast that with Rule 2 inFigure 7, a Trend Score rule. This is a trendfollowing rule based on the Directional Move-ment Index. It says that if DMI is positive andincreasing, there is a strong uptrend in place.But if DMI is decreasing, then prices are rollingover and a weak downtrend is starting.

On the other hand, if DMI is less than orequal to zero and DMI is decreasing, a strongdowntrend exists. But if DMI is increasing, thenan uptrend is starting.

This rule always fires. Of the four possiblepatterns in this rule, one will always exist andthe rule will always provide a weight to theTrend Score expert system. The normal day-to-day behavior of the indicator is continuallymeasured.

The other indicators that determine TrendScore have rules in the same form. The weights

Figure 6 Figure 7

Figure 8

Figure 9

Rule #1

If Close is a 21-day High

and VAPCT is less than zero,

WT = Down

Rule #2

If DMI is greater than zero

and DMI is increasing,

WT = Strong Up

else, if DMI not increasing,

WT = Weak Down

If DMI is less than or equal to zero

and DMI is increasing,

WT = Weak Up

else, if DMI not increasing,

WT = Strong Down

7/29/2019 aiq

5/8AUGUST 1996

AIQ Opening Be

PUTTING IT ALL TOGETHER continued . . .

Figure 10from each of the rules are combinedinto the Trend Score. The moreindicators that show strong uptrends,the higher the Trend Score. On theother hand, if all the indicators areshowing weak or strong downtrends,

then the value of the Trend Score willbe close to -100.

Using Trend Score to Confirm aStock Selection

An example of using Trend Scoreto confirm a stock selection is illus-trated in Figure 8, the Weighted ActionList for June 14. Since we are indeedbearish by this date, many traders willbe looking for stocks to short. TheWeighted Action List shows that MST,

Mercantile Stores, has a downsideExpert Rating of 100 but the stock'sparent group has a positive TrendScore of 63. Do not fight the trend.

On the other hand, IPL, IpalcoEnterprises, has an Expert Rating of 93to the downside and a group with aTrend Score of -100. The trend is yourfriend. I know which of these two Iwould consider to short.

Using Trend Score to DetermineGroup and Sector Rotation

The Daily Group Analysis reportfor June 14 is displayed in Figure 9. Inthis case, the groups are surrogateindexes, groups of stocks created withAIQs MatchMaker to represent theFidelity Select Funds.

This report is telling Select Fundtraders to transfer funds from the SelectFunds into the Money Market fund andtake a trip. According to the summaryat the top of this report, 89% of thesurrogate indexes are in a downtrend,

and 86% have a Trend Score lower thanthe prior period. Now that is bearish.

For Non-Timid Traders

The last example, Figure 10, is forthe swashbuckling buccaneers amongus. This is another Group Analysisreport for June 14 containing surrogateindex groups again constructed withMatchMaker. But in this case, theunderlying indexes all have indexoptions available. If one wants to play

trading process. As we have illus-trated, it can be used for market timinconfirmation, determining sectorrotation, analyzing groups, and pickinstocks.

If the trend is our friend, then itgoes without saying that the TrendScore is as well. n

the downward correction, a put or twoon the CBOE Environmental Indexmight be called for. This index has aTrend Score of -66 and a Delta TrendScore of -113, the fastest decliningindex on the report.

Figure 11 shows the movement ofthe CBOE Environmental Index in thesubsequent weeks. Thats what I calltrading the trend.

Trend Score is a powerful tool toadd to our kit bag of tools in our

Figure 11

7/29/2019 aiq

6/86 AUGUST 1996

AIQ Opening Bell

AIQ'S NEWMARKET BREADTH BUILDERSYSTEM TESTING

By David Vomund

Last month we introduced theMarket Breadth Builder,available in TradingExperts

new 3.0 version. The Breadth Buildercreates breadth data which allows us touse indicators such as the AdvanceDecline Line, Advance Decline Oscilla-tor, and New Highs/New Lows forgroups, sectors, or user created lists. Asecondary benefit is that we can use thevery successful AIQ market timingmodel for groups, sectors, or usercreated lists. In this article, well beginto test Expert Rating signals generatedby the market timing model for theAIQ Pyramid industry sectors.

Breadth data is most valid whencreated from a large group of stocks.The average AIQ Pyramid industrygroup has only four stocks, which isnot enough. Instead, well test theBreadth Builder on the Pyramidsectors. It will then calculate breadth

on all the stocks within each sector.We began by testing eight of the

fifteen Pyramid sectors (next monthwell test the remaining sectors). Forthese sectors, we compared the regularExpert Ratings to the market ExpertRatings which were derived using theBreadth Builder. The time period forthis test, January 1994 to July 15, 1996,covers approximately two and one-halfyears. This time period includes both asideways market and a strongly

advancing market.This is a strictly mechanical test

which considers only signals of 95 orgreater and uses the Phase indicator forconfirmation. If a buy signal of 95 orgreater is registered, we bought thesector when the Phase indicator beganto increase (or on the day of the signalif the Phase was already increasing).Conversely, we sold when a sell signalof 95 or greater was confirmed by adecreasing Phase indicator.

This is a test of the accuracy of theExpert Ratings and not a test of anactual trading strategy. After all, youcant buy a sector. Therefore, we makethe assumption that we could buy atthe close on the day of confirmation.Commissions and slippage are notincluded.

Table 1 shows the results of theeight sectors we selected. The firstcolumn shows the average annual rateof return from trading the sectors based

on the Expert Ratings from the regularAIQ stock timing model. The secondcolumn lists the results of timing thesectors using Breadth Builder to createa market for each sector. The finalcolumn lists the buy-and-hold results.

We first notice that the results withour regular knowledge base on sectorsare unimpressive. This confirms ourlong held belief that regular (stock)Expert Ratings do not work for groupsand sectors. Instead, it is better to useTrend Score and Delta Trend Scorevalues (this is why our group andsector reports are based on Trend Scoreinstead of Expert Ratings).

Next, we see that results dramati-

cally improve when we use the markettiming model on sectors instead of theregular model. On average, the returnwith the market model is more thandouble the return with the regularmodel.

Despite the fact that performancedramatically improves by creating amarket for each sector, we see that itstill does not beat a buy-and-holdstrategy for the 2-1/2 year test period.Does this mean that the new ExpertRatings should be ignored? Not sofast. Until this July, weve been in amarket that has been, for the most part,

steadily rising - a buy-and-hold typemarket. It is hard for a timer tooutperform when there are no correc-tions. Looking at the individualsectors, the sectors that outperformedthe most were those that experienced acorrection, such as was the case for theConsumer-Cyclical and Transportationsector.

A second and more importantconsideration is that the annualizedreturn of the average trade based on

the market model is greater than theannualized buy and hold return. Table2 lists the average return and theaverage holding period for each buy

Average Annual Rate of Return (%)

BreadthRegular Builder Buy &

ER Trading ER Trading Hold

Basic Materials 7.83 5.64 8.54Consumer-Cyclical 1.09 9.41 -1.14Financial Services 4.15 7.52 15.97Health & Pharm. 2.03 8.15 30.97Oil-Prod. & Services 8.46 18.37 15.27Technology -5.85 17.65 19.49Transportation 9.17 5.39 1.17Utility 4.30 0.03 2.49

Average = 3.90 9.02 11.60

Table 1

MARKET ANALYSIS

7/29/2019 aiq

7/8AUGUST 1996

AIQ Opening Be

signal. The second column annualizesthe first column. The final columnshows the annual (buy-and-hold)return for each sector.

We see that, in six of the eightcases, the annualized return for anExpert Rating buy period is greaterthan the annual buy-and-hold return.This is even true for sectors such asFinancial for which the timing modelseverely underperformed buy-and-hold (see Table 1). How can this be?The Financial sector on average made1.5% in 18 days during each buy signal.This is nearly twice the rate of returnfor this sector during an average 18 daytime period.

The reason that the timing model

underperformed is that the model wason a sell during some periods when thesector rallied. Therefore, the timingmodel outperformed when it was on abuy signal but underperformed overallbecause it didnt catch every move.This tells us that the Expert Ratingsdont catch every bullish rally butwhen the sector is on a buy signal, itoutperforms.

Summary

For the eight sectors tested, we seethat a mechanical model which buysand sells sectors based on market

Table 2

STOCK DATA MAINTENANCE

The following table shows past and future stock splits and large dividends:

Stock Ticker Split/Div. Approx. Date

Palm Harbor Homes PHHM 5:4 08/05/96

ABM Industries ABM 2:1 08/06/96

Cubic Corp CUB 3:2 08/08/96

ACC Corp ACCC 3:2 08/09/96

Bank of NY BK 2:1 08/09/96

Hubbell Inc. HUB.B 2:1 08/12/96

Werner Enterprise WERN 3:2 08/12/96

Heritage Media HTG 2:1 08/13/96

NTN Canada Inc. NTNC 3:2 08/15/96

Magna Bancorp MGNL 2:1 08/16/96

Stock Ticker Split/Div. Approx. Date

Beckton Dickinson BDX 2:1 08/16/96

Nobility Homes NOBH 3:2 08/19/96

Wolverine Worldwide WWW 3:2 08/19/96

Data Research Assoc. DRAI 3:2 08/20/96

Graham Corp GHM 3:2 08/26/96

Keystone Financial KSTN 3:2 08/26/96

Infodata Systems INFD 2:1 08/26/96

Life Tech. LTEK 3:2 08/29/96

Coachmen Inds. COA 2:1 08/29/96

SunAmerica SAI 2:1 09/02/96

Expert Ratings underperformed a buy-and-hold strategy. Thats mainlybecause of the rallies that were missed.The Expert Ratings tend to be counter-trend so if a buy signal is not regis-tered when a sector is near a low andbeginning to rally, then it is not likelythat a buy signal will occur as thesector enters its strong advance.

While the sector Expert Ratingsdont work for someone who's looking

for a black box trading system, they areeffective for a trader who wants to buyoutperforming sectors. When a sector

is on a buy signal, its average return(on an annualized basis) exceeds itsaverage buy-and-hold return. TheExpert Ratings are effective in flaggingbullish moves.

Note: The results and conclusionof our sector testing does not apply tothe overall market or individual stockWeve already seen that the markettiming Expert Ratings can effectively bapplied to timing the S&P 500 using a

mechanical trading method.n

MARKET ANALYSIS continued . . .

Trading Suspended:

Acme Cleveland (AMT) Community Health Systems (CYH)

Rockefeller Center Prop. (RCP) Ameridata Technologies (ADA)

Breadth Builder Test Results

Buy &ER Trading ER Trading Hold

% Gain per Trade Annualized % Annual %

Basic Materials 1.4 in 39 days 12.91 8.54Consumer-Cyclical 1.7 in 30 days 20.88 -1.14Financial Services 1.5 in 18 days 30.00 15.97Health & Pharm. 1.9 in 24 days 27.10 30.97Oil-Prod. & Services 3.3 in 44 days 27.50 15.27Technology 2.7 in 28 days 33.75 19.49Transportation 1.1 in 38 days 11.00 1.17Utility 0.1 in 28 days 0.13 2.49

Average = 20.41 11.60

7/29/2019 aiq

8/88 AUGUST 1996

AIQ Opening Bell

DAVID VOMUND

WHAT GOES UP FAST OFTEN GOESDOWN FASTERBy David Vomund

The month of July saw therevenge of the market timers. Ithas been years since weve seen

a serious market correction. Until July,the best rated mutual funds andinvestment newsletters were those thatremained fully invested with high betastocks. Then people learned that what

goes up fast often goes down faster.For the month, the S&P 500 fell 4.6%but the broader markets were muchweaker. The technology heavyNASDAQ Composite fell 8.8%.

The S&P 500 index fell below thesupport trendline that held the marketfor 16 months (see Figure 12). Thispattern typically means that the rate ofadvance is over and that a consolida-tion or a correction will begin. Thebearish picture was clouded, however,

as AIQ buy signals were registered asthe S&P 500 fell below the supporttrendline.

The month began with a 99 sellsignal on July 5. This was the thirdstraight sell signal with the first sellcoming on June 7. After that, there was

came a week after the turnaround day.

While the large-cap stocks haveshown better relative strength than thesmaller company stocks, that may bechanging. For the first time since May,the RSMD SPX indicator, whichcompares the Russell 2000 versus theS&P 500, is flattening (see Figure 13).This indicator will begin to advance ifthe smaller company stocks takeleadership.

Not surprisingly, volatility read-ings on the market are sky high. At thestart of 96, implied volatility readingson the S&P 500, derived from the pricesof near-term options, were runningaround 8%. Recently, volatilities haveshot up above 18%, revisiting the highslast seen in February 94.

Market timers got their revenge inthe month of July but calling a correc-tion is only half the game. A common

mistake market timers make is to fall inlove with their outlook rather thanlisten to what the market is saying.Sidestepping a correction/bear marketis no good if you end up buying backin at higher prices. Beware of becom-ing complacent. n

In addition to managed accounts,

David Vomund publishes two advisoriesfor stock and sector fund investing. For

information, phone 702-831-1544.

a whole series of buysignals. Buys wereregistered on July 9,

July 10, July 12, andJuly 23. Whilemarket commentatorswere talking aboutthe bear, AIQ signalswere saying, not sofast!

Weve reportedin the past that

historical studiesshow the highestreturn from tradingthe market timing system comes whenno confirmation is used. Adding aconfirmation does help avoid somevery painful trades, however. Here is acase where adding a confirmation wasappropriate. The first set of markettiming buy signals were not confirmedas the market was falling sharply.Confirmation of the Phase indicatorcame on July 18 when the market

began to stabilize. The last buy signalcame on the day of the Dows lowprice.

The markets intraday low pointcame on July 16, when the Dow wasdown 180 points before making adramatic turnaround and closing

higher. A bigreversal day alsocame near theend of the bearmarket in 1990

and after themini-crash of1989 (see 09/28/90 and 10/16/89).In each case, themarket made ashort term rally,fell back to thelows, and thenbegan to rally.This month, thetest of the lows

MARKET REVIEW

Figure 12

Figure 13