As of 04/18/2024
Indus: 37,775 +22.07 +0.1%
Trans: 14,947 -39.62 -0.3%
Utils: 860 +6.95 +0.8%
Nasdaq: 15,602 -81.87 -0.5%
S&P 500: 5,011 -11.09 -0.2%
|
YTD
+0.2%
-6.0%
-2.4%
+3.9%
+5.1%
|
 39,800 or 37,150 by 05/01/202416,200 or 15,000 by 05/01/2024885 or 850 by 05/01/202416,700 or 15,800 by 05/01/20245,250 or 5,025 by 05/01/2024 |
As of 04/18/2024
Indus: 37,775 +22.07 +0.1%
Trans: 14,947 -39.62 -0.3%
Utils: 860 +6.95 +0.8%
Nasdaq: 15,602 -81.87 -0.5%
S&P 500: 5,011 -11.09 -0.2%
|
YTD
+0.2%
-6.0%
-2.4%
+3.9%
+5.1%
| |
|
39,800 or 37,150 by 05/01/202416,200 or 15,000 by 05/01/2024885 or 850 by 05/01/202416,700 or 15,800 by 05/01/20245,250 or 5,025 by 05/01/2024 | ||
This article discusses the trouble with oscillators in the stock market.
Active Trader magazine's staff wrote an article titled "The trouble with oscillators" for the August 2006 issue, and I'd like to share some of their conclusions.
They compared the fast stochastic %k, commodity channel index (CCI), relative strength index (RSI), momentum (10 day close to close change), and price oscillator (close to average close 10 days ago). They paired the oscillators to check for correlation using daily data from the iShares MSCI France index fund (EWQ) from April 2001 to March 2006. I'm not sure why they chose that fund and the article doesn't explain either.
For those of you not understanding what correlation is, it's a way of answering questions like, "Do people good at math make better engineers?" or "Do people good at math make better artists?" In this case, it measures how closely the various oscillators issue the same signals near the same time.
In eight of ten tests for correlation, all but two of the oscillator combinations (Fast %k/momentum and CCI/momentum) resulted in correlations above 0.80 (highly correlated) with the other two having correlations of 0.73 and 0.70 respectively. In other words, despite the different construction of each oscillator, most of them issued the same trading signals during the period studied.
They go on to say that it's a myth that "Every oscillator is a unique indicator with unique characteristics." Oscillators use some variation of momentum and there are few ways to calculate that: price change over time or the ratio of the derivatives of price change (such as moving averages).
It's a myth that "oscillators lead price action." They write that "Oscillators cannot 'lead' price: All indicators or mathematical calculations are derivatives of price; every oscillator value depends on price action that has already occurred. Nothing can occur in an oscillator before it occurs in price."
Finally, it's a myth that "certain look-back periods or indicator settings are 'better' than others." They say that the markets are dynamic and that one setting today might be useless going forward. They say that many oscillators were developed before personal computers and so testing was error prone and not exhaustive.
For example, they say that Wilder based his 14-day RSI look back on half the 28-day lunar cycle. Translating the cycle into trading days (20 to 21 per month) means the look back should be 10-11 days and not 14.
What about oscillators with adaptive settings? "Some analysts have designed dynamic indicators that adjust oscillator parameters depending on volatility and directional changes, but such adjustments typically have limited effectiveness in addressing the root problems."
Of course, some of this analysis is based on using one obscure ETF for a 5-year period, but they make good points. If you are using multiple oscillators in your trading system, graph them on the same chart and see if they all turn near the same time. Check their correlation. Do they really add value?
-- Thomas Bulkowski
Support this site! Clicking any of the books (below) takes you to
Amazon.com If you buy ANYTHING while there, they pay for the referral.
Legal notice for paid links: "As an Amazon Associate I earn from qualifying purchases."
|
My Stock Market Books
|
My Novels
|
You'll never be the man your mother was!
on 