Tabell’s Market Letter – January 10, 1992

TABELL'S MARKET LETTER 5 VAUGHN DRIVE, CN 5209, PRINCETON, NEW JERSEY 08543-5209 MEMBER NATIONAL ASSOCIATION OF SECURITIES DEALERS, INC 1609) 987-2300 January 10, 1992 Chents (and the press) have become aware, over the years, that Delafield, Harvey, Tabell is the owner of a reasonably extensive data bank o -historical stockmarket information. -ll'fotmatjon.Tney-have aisoDCcomeaware that we actually exercise of searching through that data bank for what might be useful–or, at least, interesting information, We are, therefore, used to fielding questions that begin with When was the last time… as, for example, When was the last time the Dow advanced eleven days in a row The answer, as will be seen below, is, 'January, 1987'. The question was provoked by the rather astounding 1991-1992, yearend rally, which did, indeed, produce eleven consecutive daily rises for the DJIA, between December 18, 1991 and January 3, 1992. The exceptional vigor of this advance, even allowing for its taking place at the time of the year with the greatest upside bias, make it worth examining in some detail. The rally must be contrasted with the market climate which preceded it—a trading range almost ten months long in which the Dow held between, roughly, 2840 on the downside and a series of modestly higher highs between 3004 and 3077 on the upside On Friday, November 15th a 120-point, one-day decline suddenly erupted, and an unimpressive Monday rally was followed by yet lower prices. By mid-December the market was testing the bottom of the trading area which had contained it throughout the year, closing on December 10th at 2863.82 on the Dow, which, on the following day, went on to a somewhat frightening intra-day low of 2832.29–its lowest penetration since March. This almost classic technical test was successful—indeed eminently so. A tbree-day rally moved the indicator above the 2900 level on December 16th. Then, on the Monday of Christmas week, things really began in earoest, with an 88.IO-pomt, 3.00 percent advance, moving the DnA back above the 3000 level. Christmas Eve saw a 28-point nse and the two days after Christmas tacked on another fifty points, involving, on December 26th, a new closing high and, on the 27th, a move above 3100 for the first time in history New Year's week opened on December 30th with a 62-point, 2.01 advance and Friday saw the first close above 3200. The mild firmness of this week (WIth yesterday's trading producing a new high at 3209.53) was anti-climactic. So, when was the last time this sort of thing took place, and what new records were set We noted last week that the month of December, 1991 was the second best December in the history of the Dow, marginally bettered, jf any()ne cares, by December, 1903. What about those eleven consecutive advancing days Well, they were not a record either, but they represented a fairly mre occurrence. The record f()r consecutive advancing days by the DJIA is 13, scored on the 13 days ended January 20th, 1987. Second place goes to the .12 consecutive rises through December-12th,!970.Janumy-3rd,d992-marks-the-end-of-the-third-mstancc-()f-eloven ri,uIg days,-the ()ther tw() havlDg taken place m May, 1944 and January-February, 1955. (We have n() idea of the significance ()f the fact that f()ur ()f these five cases occurred in the December-January period.) It was, interestingly, consistency rather than explosiveness which set records. The best advance of the period, three percent, has been exceeded on 192 previous occasions. A round number that has attracted a fair degree of attention lately–since just about everyone now knows what market breadth is—is the figure ()f 1000 advances, this proportion of the 2100-odd stocks that n()w tend to trade on a given day being taken to represent a fairly g()od upside sessi()n. Fourteen ()f the ometeen days of the rally exceeded this figure as did eleven of the last thirteen. As was the case with the D()w, consIstency was the feature. The peak figure of 1410 daily advances on December 23rd had been bettered fifteen times in the past. (There were 1756 advancing issues on Oct()ber 21, 1987.) There are many ways of massaging breadth figures. (We suspect, with this week's plethora of requests, we've tned most of them.) It must first be noted that there is liUle p()mt in using a raw number such as 1000 advances. It is only recently that sufficient issues have begun to trade to allow 1000 advances to take place. One statistic that manages to relate the present to the past may be derived by l()()king for instances in which 50 ()r m()re ()f all issues traded post a rise. This occurs around the 1050-1100 level in present markets. Of the last seven trading days ()f 1991, six produced more than 50 gainers from among NYSE-traded issues. This is. since 1926, the 35th such occurrence. It is neither so rare as to be meaningless or so frequent as to be useless. It has taken place around imp()rtant bott()ms m the past, but also, less frequently, prior to important tops. Of this more later. First. however, it may be worthwhile to note an event of no significance. It has been noted by many techmcians, including ourselves, that year-end trading has abovenorma1 importance. There has emerged, however, a January Indicator, which says, 10 one version at least, that, if the first five trading days of January are up (as was, ()f course, the case in 1992), then January is likely to be up. This is true, but meaningless. Since 1897, the first five days of January have shown a rise in 61 ()f 95 cases. It is indeed true that m 43 of those 61 instances the m()nth ()f January was alS() up. H()wever, these 43 'right' calls could be made after the market had been nsing for close to a quarter of the month. It seems almost trite to observe that the predictive value of a five-day nse should be evaluated after that nse has taken place, in other words from the close of the fifth trading day to the end ()f the month rather than from the December close. When one loo. at the record in this way, it is discovered that, of the 61 times the January Indicator has predicted an up market, It has been wrong In 31. or more than 50 of all cases, Indeed a downward firstfive-trading-days is a better predictor of an up market, such a five-day penod haVIng preceded an upward remainder ()f January m 20 cases out ()f 34. Finally, what practical significance is to be read into all of thIs The record of consistently strong markets (such as markets exhibiting six of seven strong breadth days) is one which, as noted above, has characterized important market bottoms. Such techrucal strength in other words, while it may suggest a market that is overbought for the short-term is, more often than not, a precursor of intermediateterm strength. It is our Vlew that, in the current case, such should be the interpretation given the extraordinary last three weeks. It needs als() to be n()ted, th()ugh, that such strength has been–a g()od deal more rarely-ass()ciated with important tops, a phenomenon referred to in ancient texts as a blowoff. This sort of action, is, it must be stressed, relatively infrequent, but it serves to remind us that vigilance, even while maintaining optimism. IS appropriate. ANTHONY W. TABELL, CMT