Long ago, Bob Dylan sang about the need to "trust yourself" and not "be a slave to what somebody else believes." Dylan was singing about love, but that sounds like good advice for traders, too.
One day when my charting service stubbornly refused to provide daily charts, a trader advised me to substitute 720-minute charts for daily charts, noting that the moving averages would be slightly different but still deliver a workable substitute. Voila! I had a chart. From that point on, whenever my charting service balked at providing a daily chart, which it sometimes did, I switched to a 720-minute chart and went on my merry way with barely a hitch. Then, one day a reader inquired why I was substituting a 720-minute chart when each trading day was 390 minutes.
The reader was right, of course. The cash market trading day is 390 minutes. And why was I using a 720-minute chart? Someone I trusted had made the suggestion. I hadn't bothered to calculate the minutes in a trading day or test it for myself. Did it make a difference? Let's look at the charts and see.
Daily SPX Chart as of 3/24
720-minute SPX Chart as of 3/24
390-Minute SPX Chart as of 3/24:
The 390-minute and 720-minute charts prove virtually identical. Turns out that using any number of minutes over 390, the length of the trading day, produces a virtually identical chart. The longer the number of minutes used, the closer the candlesticks and moving averages appear to mimic those found on the daily chart. That proves true even if a period that's above the number of minutes in a five-day trading week (1950 minutes) is used, although logic would have told me that using a 2500-minute chart, for example, would have produced candles mimicking the weekly candles.
2500-Minute SPX Chart as of 3/24:
These examples show that although the trading day measures 390 minutes and logic suggests that a longer-period chart would grab some values from the next trading day, that just doesn't happen, at least on my charting service. Trust yourself. Try it out on your own charting service.
So where did the 720-minute calculation arise? I'm not certain. That doesn't match the hours that futures trade, either, and substitution of a 720-minute all-sessions chart for a daily all-sessions chart on the futures will not produce congruent values or candlestick formations. Because futures trade longer, they require more than 390 minutes and in fact, seem to require more than the number of minutes that futures actually trade to accurately reflect a daily chart.
Although my initial adoption of the 720-minute substitute for a daily chart
resulted in an accurate chart,
that might not have happened. Logic had suggested
otherwise, and logic suggests that traders question everything they're told by
experts of all stripes. On the morning of February 11, I benchmarked analyst
upgrades to strong buy, buy or overweight, and downgrades to sell or
I chose February 11 to benchmark upgrades and downgrades because it was a period at which I'd noted some danger signals in the forms of bearish divergences and TRAN action, hinting that markets might be nearing at least intermediate swing highs. I chose a one-month period for the test in order to allow some initial volatility after a downgrade or upgrade to subside and to move out of prime earnings season into a quieter period between earnings seasons.
First takes on the data above include the impressions that as markets were nearing their recent highs, more upgrades than downgrades were occurring; that upgraded stocks were just as likely to decline over that period as downgraded ones; and, conversely, that downgraded stocks were just as likely to gain as upgraded ones. If a trader had bought 100 shares each of those upgraded stocks and shorted 100 shares each of those downgraded ones, that trader would have lost $393.00 as of the March 10 close.
Of course, there's more to the story than that. Some upgraded stocks did swing higher in late February and early March with the broader markets.
Daily Chart of Upgraded Stock CTSH:
So did one of the four downgraded stocks.
Daily Chart of Downgraded Stock WOLF:
A closer examination that sorted gains and losses by analyst firm might have shown one firm more reliable than others. A sort by exchange might have shown differences. A look at the stocks might find that an upgrade by one firm was followed shortly by a downgrade by another or vice versa. Still, even this non-scientific, anecdotal test shows that blind trust in analysts does not always prove profitable. Look at charts. Determine whether volume patterns, particularly on-balance volume (OBV), supports the view of the analyst. Trust yourself more than the analyst.
Question everything, including the time frame someone tells you is most
reliable, the nifty new charting tool your charting service has introduced, and
analysts' comments. It's your money at stake. In the words of Dylan, "If you
want somebody you can trust, trust yourself."