Sunday, November 16, 2008

Cumulative New Highs/Lows and Stock Market Trends




One way to define a trend is a period in which more stocks are making new highs than new lows. I've taken new 20-day highs and lows across the NYSE, NASDAQ, and ASE going back to 2002 and created a cumulative line from the data. When the (pink) line is rising, we have a period in which new 20-day highs are outnumbering new 20-day lows. When the line is falling, new lows are dominating.

At times of transition, we see a handoff between new highs and lows. When the market made a momentum low in October of 2002 (top chart), the cumulative new high/low line also made a low. This was followed by a significant rally in which new highs meaningfully outnumbered new lows. When stocks retested their lows in March of 2003, the cumulative new high/low line did not confirm.

The reverse was the case at the 2007 market top. We saw new 20-day highs consistently outnumber new lows into the momentum peak at mid year (middle chart), then a topping and meaningful pullback of cumulative new highs/lows, followed by price highs in October that were not confirmed by the cumulative high/low line.

As we struggle to hold recent lows in the S&P 500 Index (bottom chart, blue line), the cumulative new high/low line continues to march to new lows. So far, we see no non-confirmations on the order of the two trend reversals illustrated above.
.

6 comments:

catkevin said...

Very clear analysis and plots on where we are now versus past data - thank you.

Kevin

David L. Spurr said...

Interesting analysis.....A sharp strong, bold, broad rally would quickly move your indicator off the bottom. The markets are extremely oversold at this point and my bet is that we're due to rally at some point whether it's today,tomorrow, next week, next month......It's hard to want to take new short positions even though that's what the indicators are suggesting

EnglishTeach said...

(Off topic)

I'm curious about your reasons for moving to the ES futures. I've traded for approximately 3 years (securities and options) and have lately been studying the futures markets. My primary job is as a college instructor, so I have quite a bit of time off during the year (Christmas/semester break and summers).

abel said...

Excellent charts, and comparisons. My initial sense, however, is the period today may have other market periods that 'correlate' more in tandem. For example, the 1973-1974 period may apply. In fact, the composition, or blending of two or more periods may be appropriate, for example, inclusion of a market bottoming period that closed a chapter of an inflationary, reasonably high-growth wave.

If looking and relying predominantly upon the charts, and also upon my interpretation, I believe there is sound basis for proposing a market bottom may be present right here. Using your charts as examples, I would propose the possibility, perhaps even the probability, that the 'price component' of the data, has exceeded the bounds of the cumulative NHNL data, by a margin wide-enough to be considered 'stretched', and possibly unsustainable, from which a rebound may occur.

It can be clearly seen that there are precedents for this dynamic, in the charts you have displayed, notably, the middle chart, where the price exceeded above the NHNL line (OCT 07), and NHNL's provided non-confirmations. I believe the same relationship exists today, in inverse, where the NHNL's are exhibiting relative strength in here, by providing non-confirmations of price lows.

Reviewing other additional periods in similar detail may provide useful precedents for the scenario above.

If I had to sum it up, for me, that's how 'it feels.'

While a move up, were it to occur, may only represent an intermediate bull move, as opposed to a more sustainable bull leg, and there is ample evidence that fundamental factors may continue to erode, providing a reasonable argument why values might decline for a continued duration, I still sense a move up may manifest from this area, in price and time.

These are my opinions only, and traders / investors will have to make their own judgements.

As I wrote this in the morning today, and wondered whether to submit it for possible posting, I would like to add that, this evening, in the Front Range of Colorado, where I have resided for ten years, the skies are presenting a beautiful display of pink and red cloud bands, against a backdrop of clear and full-hued light-to-deepening blues. The ambient light produced by this splendid combination, is a glowing rose, that is an equal match to the light found at the end of a rainbow, in its own right. The best such instance of many I've seen in those ten years.

Brett Steenbarger, Ph.D. said...

Hi English Teach,

I routinely follow futures as well as ETFs; I find the volume trends and patterns in ES futures to be helpful for my analytics--

Brett

Brett Steenbarger, Ph.D. said...

Thanks David and Abel,

I appreciate the analyses. I just can't find signs of institutions putting money to work on the buy side. No doubt it will happen at some point, but a lot of trading accounts have been shattered crystal-balling that turnaround.

Brett