You wake up in the morning, turn on the radio and the announcer says:
“…and now to the markets overnight, where the US share index fell 100 points - looks like being a black day for the Australian market today”.
If you're lucky you might hear the actual index value that the US drop relates to. If it’s the Dow, it’s currently around 24,000, so 100 points is only 0.4%. Occasionally the voice might compute the percentage for you as your early morning brain clicks into gear.
How Australia reacts in the day ahead
I studied daily one-day closing price moves on the US S&P500 index and the Australian All Ordinaries index the next day. (At this time of year, the New York market closes 8am Sydney time. The ASX closes at 4pm). This gave 6,882 days of data in the 35-year period from January 1985 to February 2020.
I excluded Monday trading in Australia because of the longer time lag from US Friday close to when Australian trading closes on the Monday. Of course, on all Australian trading days there could well be Asian market events breaking which were not reflected in US markets overnight, but the time gap is much shorter.
My analysis of the results is set out in the table below.
Table 1: Total period – 3 January 1985 to 28 February 2020
Tuesdays to Fridays of All Ords trading
|
S&P500 move +/- % bigger than this
|
% of overnights S&P500 does this
|
% of next days that All Ords ends in same direction
|
|
0.0%
|
100%
|
65%
|
|
0.2%
|
76%
|
69%
|
|
0.5%
|
50%
|
75%
|
|
1.0%
|
24%
|
82%
|
|
1.5%
|
12%
|
86%
|
|
2.0%
|
6%
|
88%
|
|
2.5%
|
4%
|
90%
|
|
5.0%
|
0.5%
|
91%
|
For moves of plus or minus 0.5% or less in the S&P500, there is no reliable pattern on which to base a strong expectation for direction of the next Australian All Ords move – i.e. the Australian market is just as likely to go in the opposite direction to the US about one day in four.
But I found a neat rule of thumb we could pass on to the finance journalists to encourage them to talk percentages while we are pouring milk on the muesli.
When the S&P500 overnight moves more than 2.5% up or down, there is a 90% likelihood the Australian market will move in the same direction. Moves above 2.5% have occurred in 4% of days over the long-term period studied – that is an average of about one trading day a month.
Variations by trading day
It is also interesting to explore variation in these results by trading day and whether there was any change over time in this 35-year period.
Firstly, let us look at trading days – and remember as explained earlier, “I don’t like Mondays”. The following table splits the results for Australian trading days for Tuesdays to Fridays.
Table 2: Days of the week variation (Australian days)
Percentage of days that the All Ords move on close is in the same direction as S&P500 close overnight
|
S&P 500 move +/- % bigger than this
|
Tuesday
|
Wednesday
|
Thursday
|
Friday
|
|
0.0%
|
64%
|
66%
|
64%
|
67%
|
|
0.2%
|
68%
|
70%
|
67%
|
71%
|
|
0.5%
|
74%
|
74%
|
74%
|
78%
|
|
1.0%
|
82%
|
80%
|
82%
|
84%
|
|
1.5%
|
86%
|
85%
|
84%
|
88%
|
|
2.0%
|
84%
|
86%
|
92%
|
92%
|
|
2.5%
|
86%
|
88%
|
94%
|
94%
|
|
5.0%
|
91%
|
89%
|
100%
|
75%
|
There is not much of a trend to observe here. The bigger than 2.5% results are slightly lower on Tuesdays and higher Thursdays and Fridays. Perhaps there is a volatility momentum effect through the week when big moves are occurring.
To look at the variation effect over time, I split the analysis into last century and this century, see table 3.
Table 3: 20th Century vs 21st Century
% of next days that the All Ords ends in same direction as S&P500 overnight
|
S&P 500 move +/- % bigger than this
|
20th Century
3 Jan 1985 to
29 Dec 2000
|
21st Century
3 Jan 2001 to
29 Feb 2020
|
21st Century excess over 20th Century
|
|
0.0%
|
62%
|
68%
|
6%
|
|
0.2%
|
65%
|
72%
|
7%
|
|
0.5%
|
71%
|
78%
|
7%
|
|
1.0%
|
79%
|
84%
|
5%
|
|
1.5%
|
81%
|
89%
|
8%
|
|
2.0%
|
86%
|
89%
|
3%
|
|
2.5%
|
90%
|
90%
|
-
|
|
5.0%
|
88%
|
92%
|
4%
|
There is a stronger trend here for following the S&P500 but mainly where there are more data points. Perhaps we could conclude that globalisation and the internet bring a stronger influence in the 21st Century.
Now before you all go off and start day trading with these results, I should add my disclaimer – don’t try this at home! The rule of thumb presented here is for general education purposes so that market commentary is encouraged to talk and think percentages and to give big market moves rational historical context.
Postscript
This article was finalised after 4pm Tuesday 10th March 2020, a time when we saw a classic exception reflected in the ‘rule of thumb’! The S&P500 was down 7.6% the night before, and the All Ords was up 3% at close the next day due to various stimulus packages announced around the world. Events can overtake overnight movements. This opposite direction has happened 10% of the time when the S&P500 moves more than 2.5%.
Bruce Gregor is a Demographer and Actuary, and Founder of Financial Demographics. This article is general information only and does not consider the circumstances of any investor.