The Timing of the QBO Reversal

[donsutherland1]

Given its significant implications for the upcoming winter, the timing of the QBO's reversal is of great interest.

Several points are in order:

Unlike in some years in which the initial effort at reversal failed and the QBO saw its East phase strengthen, this will not likely occur this year. In every year in which the initial reversal failed, the initial reversal was less than +1.50:

1956: +0.50
1962: +0.24
1965: +0.59
1974: +0.40
1986: +0.61
1991: +0.73
2000: +1.48

The initial attempt at reversal which occurred in September 2003 was +2.13.

Moreover, in every year in which the initial reversal failed when the QBO Index was -20 or less, the initial reversal was less than +1.00. Therefore, confidence is high that a reveral that will lead to a positive (West) QBO is underway.

When that transition will be completed is the next question. The answer lies with the October QBO Index (which is not yet available as the month has not yet concluded).

Excluding 2003, there have been 15 occurrences in which the QBO has fallen below -20: 1959, 1963, 1965, 1968, 1970, 1972, 1974, 1977, 1984, 1987, 1989, 1994, 1996, 1998, 2001.

In those cases, the following has held true 14 of 15 times (93%) which is compelling even when one consids the relatively small sample size:

After the QBO Index reached its lowest reading (Month 1=first month of reversal/attempt at reversal), the number of months required for it to complete its shift to a postive (West) reading came was as follows:

<b>3 Months or Less:</b>

The month 2 rise was greater than or equal to 200% of the Month 1 rise and after 2 months, the reversal was 90% or greater complete

<b>4-5 Months:</b>

The month 2 rise was greater than or equal to 200% of the month 1 rise and the reversal after two months was less than 90% complete

The month 2 rise was greater than 100% but less than 200% of the month 1 rise and the reversal after two months was 50% or more complete

<b>6 Months or More:</b>

The month 2 rise was greater than 100% but less than 200% of the month 1 rise and the reversal after two months was less than 50% complete.

The month 2 rise was less than 100% of the month 1 rise.

Examples:

1959: -20.06 -17.24 -14.02:

M2/M1=114%; 6.04/20.06=30.1%; Expected time to West QBO: 6 months or more; Took 6 months or more

1963: -24.18 -20.08 -9.27:
M2/M1=264% 14.91/24.18=61.7%; Expected time to West QBO: 4-5 months; Took 4 months

[JCT777]

Thanks for the info, Don. Very interesting!

[donsutherland1]

Unfortunately, when it comes to the timing of a shift from West to East (positive to negative), there is no definitive pattern. Indeed, the Westerly QBO weakens only to strengthen again before it finally makes a turn for negative.

I have an idea as to why this might be the case, but unfortunately, this does not allow for greater accuracy in timing West-to-East shifts.

What does seem apparent is that West-to-East shifts, on average, take longer than East-to-West shifts (the kind described at the beginning of this thread).