Some commentary on the Dow 2019

[tomorton]

We're almost at the year's end and if you spreadbet the Dow, its been pretty good, though a Twitter-driven roller-coaster at times.
Overall performance = +4800pts (+20%)

Good, but what about the drawdowns? These were bad - the 200EMA was breached in May, August and October, the first two episodes were well over 1500pts deep, the third over a thousand. But if you bought on last New Year's Eve and held, you were only underwater for 24 hours in 2019.

But nobody takes a spreadbet for a year at a time. So how good was the year in detail? In particular I wanted to see whether it was better to hold or close a winning long after strong daily closes.

The year was in detail not one long smooth uptrend. Taking a long-term uptrend to be confirmed when the 20 and 50EMA's are sloping upwards and the 20 is above the 50, that only makes just over half of the year.

Days with price increases from the prior close made up only just over half the total. Up-days in uptrends were no more common.

The majority of the year’s total positive price performance could have been obtained from just the uptrend periods, 95% (though that doesn't mean only 5% could have been obtained from ranges and downtrends!).

Half of all days with a higher close than the prior day were followed by a lower close the next day.

But if you held after 2 higher closes, most of the time the third day would have seen another higher close.

Holding a long in an uptrend beyond the third consecutive higher close would have been only a little better than a coin-toss.

Holding beyond a fourth consecutive higher close wouldn't have been possible very often, and mostly would have been an unprofitable decision.

Looking shortly at single-day price changes......


Good luck all for 2020!

 

[tomorton]

Single-day price changes.

There were 42 days so far in 2019 when the Dow rose or fell by 1% or more from the prior close.

As expected, most were bullish, 57%.

Very few large price falls came out of standard uptrends (22%). Likewise, very few large price rises came out of downtrends (25%). These rates are close together, giving some added credibility and strengthening a case for trend-following.

 

[kalott]

Trump needs to step it up a notch

Stock Market Performance by President

This interactive chart shows the running percentage gain in the Dow Jones Industrial Average by Presidential term. Each series begins with the closing value of the month of inauguration and runs to the closing value of the last month of the term. The y-axis shows the total percentage increase...

www.macrotrends.net

😎

 

[tomorton]

Trump needs to step it up a notch

Stock Market Performance by President

This interactive chart shows the running percentage gain in the Dow Jones Industrial Average by Presidential term. Each series begins with the closing value of the month of inauguration and runs to the closing value of the last month of the term. The y-axis shows the total percentage increase...

www.macrotrends.net

😎

Love this chart, thanks kalott.

 

[tomorton]

Many traders advise caution when trading on Mondays and Fridays, or they avoid them altogether. But this is mostly due to poor results from intra-day trading. So how were these days for long-term trading the Dow?

In 2019, we know that 56% of days saw a price rise on the prior close. The Mondays in 2019 saw just the same win rate, 56%. But the Fridays performed better - 69% showed a price rise on the prior close.

Tentatively this suggests that holding from Friday through the weekend remains a matter of choice, but holding all the way to the Friday close was a tactic worth the extended exposure and trade monitoring time.

 

[tomorton]

More on the subject of Mondays and Fridays and whether they are risky days to hold. Accepting that conclusions from 1 year reflect a small sample base, just 50 of each day of the week.

However, between a third and a half of the larger single day price rises, whether or not in uptrends, occurred on a Friday during 2019. No day of the week (not even Fridays) stood out as heavily prone to larger single day price falls. Tentatively, this further suggests Friday might be a bad day to scalp but a good day to hold until the weekend. The absence of large numbers of larger Monday price falls is also contrary to the thinking that holding over the weekend is any more risky than holding over any single week-night.

 

[tomorton]

Dow holding periods -

Obviously, holding the Dow for the whole year would have produced a respectable 20% price rise. But no trader buys the Dow intending to hold for a year. In any case, the draw-downs would have been horrendous (though the worst only nudged -7%).

Is there an optimum period to hold, at the end of which you could sell at a high close, and expect to see a lower close the following day?

Holding for 1 day would have lead to a higher close just over half of the time. Holding for longer leads to an increasing probability of a lower close, skimming off potential profit levels. It looks like adjusting the chart to the uncommon 2-day time-frame produces 120 bars, but increases the bullish proportion to 63% from 56. The ratio falls back to 59% if you take the time-frame to 3-days.

So it looks like holding for 2 or 3 days would usually be optimum. This does not tally at all with the usual adage let your winners run. From notes above it looks like Wednesday to Friday or Thursday plus Friday might be the best days to be long: this again doesn't tally with what we normally hear about being in trends. Well, we'll see.

 

[tomorton]

So let's suppose you're long on the Dow but its pulling back - how long will this go on? Or you're in cash and waiting to get long - when will the pull-back end?

In 2019 we've had 99 days put of 244 on which there was a lower close than the prior day - 41%.

But if you held through the first pull-back day, 72% of the time this would have been followed by a day with a higher close. So the pull-back was only a 1-day affair.

19% of pull-backs were 2-days only with lower consecutive closes, 5% were 3, 1% were 4 and 2% were 5. So on 92% of occasions, the pull-back would only run to 2 lower closes at worst and on only 3% of occasions did a 1-day pull-back develop into a more than 3-day pull-back.