Research & Advice

|

Disruption of the Primary Trend

Sunday, May 16, 2010

 

  • There have been no typical signs of a major top in the stock market. The current correction so far appears to be nothing more than a disruption to the primary bull market trend.

  • While a price low has likely been made for this correction, that low is likely to be tested one or more times over the coming weeks.

  • Just as the run up in stock prices during April was an opportunity to reduce equity positions (as Berkshire Money Management did), this current pullback is an opportunity to increase equity positions.

  • The magnitude and duration of the bull market from the March 2009 low has been impressive, but remains consistent with previous bull markets.

 

As any experienced investor knows, anything is possible in the equity market. But at Berkshire Money Management we deal more with probabilities than possibilities. While it is possible that any deviation from short-term stock market highs could devolve into a new bear market, it is most probable that the current correction is nothing more than a disruption of the primary trend. And the primary trend continues to be a cyclical bull market. Not only were there few of the historical indications of a major market top in place prior to the April highs, but the recent pullback itself has so far appeared more characteristic of a market correction than that of a decline that initiates the resumption of a new bear market.

In terms of the of the current market decline, the drop in the Dow Jones Industrial Average (DJIA) and the S&P 500 from the April highs to the early May lows was the most severe of any pullback since the March 2009 market bottom. Using the DJIA, from April 26th to May 7th, a span of ten trading days, the market dropped 7.4% on a closing basis. In the January-February 2010 decline, the DJIA lost 7.6%, but over fourteen days. And in the June-July 2009 correction the DJIA also dropped 7.4%, but over nineteen days. The magnitude of the current pullback is in keeping not only with this bull market, but with past bull markets.

Such a swift drop, coupled with such a notable rise in volatility, naturally begs the question if this is the start of a new bear market. In regards to investing, it is a dangerous game to utter the phrase “this time it’s different”. And while it is possible that sometimes, albeit rarely, it is actually different, the odds favor this pullback being nothing more than a painful correction that will be resolved with higher highs in the stock market (even if later, rather than sooner).

Typical signs of an impending bear market are not present. For example, at significant market tops there are non-confirmations of price highs for indices by advance-decline lines and the number of new highs for stocks. That’s industry jargon for saying that a stock market’s continued advance is suspect if it continues to go higher based on fewer and fewer individual stock prices propping up the market as a whole.

At major market tops more and more stocks drop out of their advancing trends. Currently the opposite pattern exists (or at least ‘existed’, as corrections temporary alter that pattern) as buying enthusiasm has been broadening, rather than contracting. This change in selectivity typically occurs months prior to the top of market indexes, which would suggest that a market top is, at least, months away.

Furthermore, the “flash crash” of May 6th and the subsequent May 7th drop have temporarily exhausted Supply (i.e. selling). The following Monday and Wednesday (May 10th and 12th) saw robust Demand (i.e. buying) which strongly suggests a price level at which investors are willing to buy stocks not just for a trade, but for longer-term accumulation. This is very typical price action at the lows for corrections.

The intensity of the May 6th and 7th declines seems characteristic of capitulation selling. Typically, surges in volume will occur around market turning points (at both tops and bottoms), and both of those days certainly qualified as displaying a surge in volume. The reason this surge occurs around turning points is because it frequently represents an exhaustion of Supply (i.e. selling) at market bottoms and an exhaustion of Demand (i.e. buying) at market tops. In a true selling climax, the exhaustion in Supply is replaced by strong Demand (such as occurred on the 10th and 12th). This shift in control from sellers to buyers usually manifest itself by an intraday rally that produces a close near the high for the day (again, such as occurred on the 10th and 12th).

Taken together, the days of exhaustive selling followed by the rejuvenated buying as well as the lack of price divergence at the market’s top are much more consistent with a normal correction in an ongoing primary uptrend than with the beginning of a new bear market.

However, the probabilities do strongly favor an eventual test of the May 6th and 7th lows (the intraday low of May 6th was 1,065 points on the S&P 500, but that low was suspect – our April 4th, 2010 prediction of 1,085 points still seems the most reasonable). Historically, selling climaxes and deeply oversold readings are more likely to be tested one or more times than sell-offs that end less violently.

In the “Second Quarter Update for the Economic Outlook for 2010” we did predict a longer than typical correction that would bring the S&P 500 down to 1,085 points. This one has, so far, been quicker than the previous corrections. And the closing low has so far only been 1,100 (as of May 7th). Getting to 1,085 on a closing basis is not a given, but it is probable that the stock market will retest its May 7 low and it is also likely for that test to last weeks, not days. The likelihood that this current pullback is merely a disruption of the primary trend does not at all negate the probability for one or more tests of the recent low.

If the 1,085-point level is violated for too long, we reserve the right to change our mind regarding our more optimistic tone and to react by selling equity and raising cash. Currently the 200-day moving average is roughly 1,100 points. We have in the past written about the use of technicals to reduce risk. Perhaps we will revisit those instructionals again in the future, but for now let us briefly acknowledge that the trend is often (but not always) more important than the research (i.e. probabilities).

As John Maynard Keynes once said, “markets can remain irrational longer than you can remain solvent”. So while we are confident that our expectations and forecasts are completely sound and based on world-class research (and are thus rational), we also understand that the markets do not have to accept our homework, no matter how good it may be for any particular situation.

Just as Berkshire Money Management viewed the run-up into April an opportunity to reduce equity positions (which we did), we now view this pullback as a buying opportunity. While we at Berkshire Money Management enjoy taking risks (and that is exactly what investing is – taking a risk) and reaping the rewards of such actions, we are simultaneously conservative. An oxymoron as it may be, the conservative risk-taker can make any trade (investment) they wish to so long as they have an exit plan.

Berkshire Money Management’s exit plan typically is to identify stock market tops and then to work down equity positions (just as we did in 2001-2002 and 2007-2009). But if we build up equity positions in the expectation of a rebound that never comes, our exit plan is to reduce equity/raise cash if we continue to cross and hold below certain technical levels (such as a declining 200-day moving average).

But what of the fundamentals? On June 30, 2009 we wrote that the recession would be over by the end of that summer and that the recovery would begin. With nearly a years’ worth of hindsight and revised data available it is safe to say not only was that a correct prediction, but that the recovery has been much stronger than nearly anyone had expected. When the S&P 500 dropped below the 1,100 point level in the fall of 2008, the trailing 12-months earnings for companies that comprised that index (using “As Reported” numbers, which are very close to Generally Accepted Accounting Principles – there are actually different types of “earnings”) was $14.88 (even if you ignore the negative fourth quarter of 2008 and its massive financial company write offs, the earnings were still only $38.13). Currently the trailing 12-months of earnings are $48.58, thus more supportive of current prices.

Instead of contracting, as it was in the fall of 2008, the economy is now growing. Further, instead of failed legislation and failed banks in the fall of 2008, we now have $787 billion worth of stimulus spending and mergers & acquisitions. Instead of bank runs, we now have expanding credit. Instead of losing a half million jobs per month, a half million jobs were created over the last two months. Instead of a temporary cultural shift of shunning excess, we now have Apple selling one million iPads even before consumers fully know what the product is.

We cannot yet say that fear has been replaced with greed, but the landscape has changed considerably for the positive. The fundamental environment doesn’t get much better for equities – earnings growth is consistently strong globally and the improved economic momentum supports the prospects for more of the same. Inflation is and remains low. Interest rates are absurdly low. It is a powerful mix for supporting stock prices.

But what of the concern that this rally has ran too far too soon? On the last pages of this report is a statistical package for you to review. The package tracks the movement of the DJIA and allows us to compare this rally to past bull markets. This rally’s gain of 71.1% in the DJIA ranks 18th among the 34 cyclical bulls since 1900, and its 410-day duration ranks 23rd.

Like a broken record, over nearly a decade of doing business as Berkshire Money Management we have been saying that we are in a secular (i.e. long-term) bear market. This rally, we believe, is a cyclical bull within a secular bear. Among the 16 cyclical bulls that have taken place within secular bears, this gain ranks only 8th and the longevity 9th. While we have not yet seen a drop of more than ten percent since the low of March 2009, 17 other advances lasted longer without one. This rally, while impressive, is not a historical aberration. This rally has so far been fairly consistent with previous bull markets. Maybe, once again, it is not different this time.

 

 

Bottom Line: It is most probable that the current correction is nothing more than a disruption of the primary trend. And the primary trend continues to be a cyclical bull market. Not only were there few of the historical indications of a major market top in place prior the April highs, but the recent pullback itself has so far appeared more characteristic of a market correction than that of a decline that initiates the resumption of a new bear market.

It is often said that a sell-off is a “healthy correction” if it corrects an overbought condition and relieves excessive optimism, doing so without imperiling the longer-term uptrend. The April-May correction fits that description. Not only did it produce an oversold condition and bring sentiment indicators back down to more favorable levels, but it occurred with dropping bond yields. Reassured by the $1 trillion loan package in Europe, the markets can turn their attention back to the improved global economic environment, earnings growth, and low interest rates – all conditions that support the continuation of the cyclical bull market.

The likelihood that this current pullback is merely a disruption of the primary trend does not at all negate the probability for one or more tests of the recent low. Just as Berkshire Money Management viewed the run-up into April an opportunity to reduce equity positions (which we did), we view this pullback as a buying opportunity.

 

 

 

 

 

 

  • Run Date = Fri May 14 18:56:25 2010
  • Cyclical bull and bear markets within secular bear markets highlighted in gray and italicized.

Cyclical Bull Markets

Cyclical Bear Markets

Start Date

DJIA

End Date

DJIA

% Gain

% GPA

Days

09/24/1900

38.80

06/17/1901

57.33

47.8

70.9

266

11/09/1903

30.88

01/19/1906

75.45

144.4

50.2

802

11/15/1907

38.83

11/19/1909

73.64

89.7

37.4

735

09/25/1911

53.43

09/30/1912

68.97

29.1

28.5

371

12/24/1914

53.17

11/21/1916

110.15

107.2

46.4

698

12/19/1917

65.95

11/03/1919

119.62

81.4

37.4

684

08/24/1921

63.90

03/20/1923

105.38

64.9

37.5

573

10/27/1923

85.76

09/03/1929

381.17

344.5

29.0

2138

11/13/1929

198.69

04/17/1930

294.07

48.0

151.8

155

07/08/1932

41.22

09/07/1932

79.93

93.9

5158.9

61

02/27/1933

50.16

02/05/1934

110.74

120.8

132.3

343

07/26/1934

85.51

03/10/1937

194.40

127.3

36.7

958

03/31/1938

98.95

11/12/1938

158.41

60.1

113.8

226

04/08/1939

121.44

09/12/1939

155.92

28.4

78.8

157

04/28/1942

92.92

05/29/1946

212.50

128.7

22.4

1492

05/17/1947

163.21

06/15/1948

193.16

18.4

16.8

395

06/13/1949

161.60

01/05/1953

293.79

81.8

18.2

1302

09/14/1953

255.49

04/06/1956

521.05

103.9

32.1

935

10/22/1957

419.79

01/05/1960

685.47

63.3

24.9

805

10/25/1960

566.05

12/13/1961

734.91

29.8

25.9

414

06/26/1962

535.76

02/09/1966

995.15

85.7

18.6

1324

10/07/1966

744.32

12/03/1968

985.21

32.4

13.9

788

05/26/1970

631.16

04/28/1971

950.82

50.6

55.9

337

11/23/1971

797.97

01/11/1973

1051.70

31.8

27.5

415

12/06/1974

577.60

09/21/1976

1014.79

75.7

36.9

655

02/28/1978

742.12

09/08/1978

907.74

22.3

46.7

192

04/21/1980

759.13

04/27/1981

1024.05

34.9

34.2

371

08/12/1982

776.92

11/29/1983

1287.20

65.7

47.5

474

07/24/1984

1086.57

08/25/1987

2722.42

150.6

34.6

1127

10/19/1987

1738.74

07/16/1990

2999.75

72.5

22.0

1001

10/11/1990

2365.10

07/17/1998

9337.97

294.8

19.3

2836

08/31/1998

7539.07

01/14/2000

11722.98

55.5

37.9

501

09/21/2001

8235.81

03/19/2002

10635.25

29.1

68.4

179

10/09/2002

7286.27

10/09/2007

14164.53

94.4

14.2

1826

03/09/2009

6547.05

*4/26/2010

11205.03

71.1

60.8

413

Start Date

DJIA

End Date

DJIA

% Gain

% GPA

Days

06/17/1901

57.33

11/09/1903

30.88

-46.1

-22.8

875

01/19/1906

75.45

11/15/1907

38.83

-48.5

-30.6

665

11/19/1909

73.64

09/25/1911

53.43

-27.4

-15.9

675

09/30/1912

68.97

07/30/1914

52.32

-24.1

-14.0

668

11/21/1916

110.15

12/19/1917

65.95

-40.1

-37.9

393

11/03/1919

119.62

08/24/1921

63.90

-46.6

-29.3

660

03/20/1923

105.38

10/27/1923

85.76

-18.6

-28.8

221

09/03/1929

381.17

11/13/1929

198.69

-47.9

-96.5

71

04/17/1930

294.07

07/08/1932

41.22

-86.0

-58.6

813

09/07/1932

79.93

02/27/1933

50.16

-37.2

-62.6

173

02/05/1934

110.74

07/26/1934

85.51

-22.8

-42.4

171

03/10/1937

194.40

03/31/1938

98.95

-49.1

-47.2

386

11/12/1938

158.41

04/08/1939

121.44

-23.3

-48.3

147

09/12/1939

155.92

04/28/1942

92.92

-40.4

-17.9

959

05/29/1946

212.50

05/17/1947

163.21

-23.2

-23.9

353

06/15/1948

193.16

06/13/1949

161.60

-16.3

-16.4

363

01/05/1953

293.79

09/14/1953

255.49

-13.0

-18.3

252

04/06/1956

521.05

10/22/1957

419.79

-19.4

-13.1

564

01/05/1960

685.47

10/25/1960

566.05

-17.4

-21.2

294

12/13/1961

734.91

06/26/1962

535.76

-27.1

-44.7

195

02/09/1966

995.15

10/07/1966

744.32

-25.2

-35.7

240

12/03/1968

985.21

05/26/1970

631.16

-35.9

-26.0

539

04/28/1971

950.82

11/23/1971

797.97

-16.1

-26.4

209

01/11/1973

1051.70

12/06/1974

577.60

-45.1

-27.0

694

09/21/1976

1014.79

02/28/1978

742.12

-26.9

-19.6

525

09/08/1978

907.74

04/21/1980

759.13

-16.4

-10.5

591

04/27/1981

1024.05

08/12/1982

776.92

-24.1

-19.2

472

11/29/1983

1287.20

07/24/1984

1086.57

-15.6

-22.9

238

08/25/1987

2722.42

10/19/1987

1738.74

-36.1

-94.9

55

07/16/1990

2999.75

10/11/1990

2365.10

-21.2

-63.1

87

07/17/1998

9337.97

08/31/1998

7539.07

-19.3

-82.4

45

01/14/2000

11722.98

09/21/2001

8235.81

-29.7

-18.9

616

03/19/2002

10635.25

10/09/2002

7286.27

-31.5

-49.2

204

10/09/2007

14164.53

03/09/2009

6547.05

-53.8

-42.0

517

 

All Cyclical Bull Markets

All Cyclical Bear Markets

Statistic

# Cases

% Gain

% GPA

Days

Mean

34

85.6

194.9

751

Median

34

69.1

36.8

614

Statistic

# Cases

% Gain

% GPA

Days

Mean

34

-31.5

-36.1

410

Median

34

-27.0

-27.9

375

 

Cyclical Bulls

Within Secular Bulls

Cyclical Bears

Within Secular Bulls

Statistic

# Cases

% Gain

% GPA

Days

Mean

16

109.5

31.7

1024

Median

16

77.2

27.4

870

Statistic

# Cases

% Gain

% GPA

Days

Mean

12

-22.8

-37.7

295

Median

12

-19.4

-23.4

245

 

Cyclical Bulls

Within Secular Bears

Cyclical Bears

Within Secular Bears

Statistic

# Cases

% Gain

% GPA

Days

Mean

18

64.3

340.0

508

Median

18

55.4

41.9

371

Statistic

# Cases

% Gain

% GPA

Days

Mean

22

-36.3

-35.3

472

Median

22

-33.7

-30.0

521

 

  1. A secular bull market is a period in which stock prices rise at an above-average rate for an extended period and suffer

    only relatively short intervening declines. A secular bear market is an extended period of flat or declining stock prices.

    Secular bull or bear markets typically consist of multiple cyclical bull and bear markets.

  2. A cyclical bull market requires a 30% rise in the DJIA after 50 calendar days or a 13% rise after 155 calendar days. Reversals

    of 30% in the Value Line Geometric Index since 1965 also qualify. A bear cyclical market requires a 30% drop in the DJIA after 50

    calendar days or a 13% decline after 145 calendar days. Reversals in the Value Line Geometric Index also qualify.

  3. *The current cyclical bull/bear market is bolded. The corresponding end date is the date of the high/low from the

    trough or peak. As such, the current bull/bear stats could change with a new price peak or trough.

  4. Summary stats do not include the current cyclical bull or bear market.
  5. GPA = gain per annum. For bull and bear markets lasting less than one year, the percent GPA will be greater than the

    percent gain. Extremely high (low) GPAs can skew the mean statistics. As a result, the mean GPAs can be much higher (lower)

    than the median GPAs.

  6. Days = calendar days