RSI and the Old Skeptic

[Dividend Watcher]

In the thread about the short dip in early 2014 here, Be Here Now added this comment about the Relative Strength Index.

(02-15-2014, 08:03 PM)Be Here Now Wrote: I used the RSI (relative strength index) technical indicator to identify, as close as one can in these things as they happen, the best time to buy both. For those who are not familiar with the RSI, readings below 20 are short term highly oversold, and readings above 80 are short term highly overbought.

I don't use technical indicators. I've always been turned off by those graphs with squiggles and lines drawn on them. Half the time they seem drawn arbitrarily, not consistently connected to any points or events on a chart, just to show some pattern or movement.

So, after Bob Johnson on SA mentioning it and now BHN mentioning it here, I ambled over to Investopedia and read a little more about it. Basically, it's a ratio of up vs. down closings for the last 14 days according to the article I read. The full formula is located here on the Investopedia web site.

To me it seems a short-term indicator. If I wanted to do a lot of trading, which I don't, this may be one tool in the toolbox.

Where I see it could be useful is a confirmation that it's a good time to buy when I'm looking at initiating or adding to a position. Since I'm not inclined to sell much, I don't think it would have as much an influence on my sell decision. Usually I only sell to improve the portfolio or there's too much uncertainty around a particular company.

So, my questions are:

1.) Do you use RSI in making buy & sell decisions and has it been a fairly reliable indicator?

2.) What about other levels? Suppose you were interested in the stock and it met your other criteria. Would you hold off buying if the RSI was 40? Does the trend tell you anything?

I haven't looked at too many charts with the RSI displayed so I don't have a feel for it yet.

[Dexter]

I still use the RSI with weekly/monthly charts but it is not the sole deciding factor for me when it comes time to make an investment. In fact, it is usually the only indicator I use when looking at charts.

Yes it is very reliable but things can stay overbought and oversold for a very long time.

[NilesMike]

(02-16-2014, 07:36 PM)Dexter Wrote: I still use the RSI with weekly/monthly charts but it is not the sole deciding factor for me when it comes time to make an investment. In fact, it is usually the only indicator I use when looking at charts.

Yes it is very reliable but things can stay overbought and oversold for a very long time.

Is that like saying it's right except when it's wrong?

RSI does not work to indicate what future prices will do. It can tell what past prices have done but that is clear on the charts already.

I have not seen any study or backtest that indicates otherwise regarding RSI or any other oscillator indicator.

[Dexter]

Is that like saying it's right except when it's wrong?


[/quote]

Yes, that's exactly what I'm saying

[hendi_alex]

I find MACD and RSI to be useful in timing purchases. IMO it would be ridiculous to use them in isolation to other factors, but they can throw the odds in one's favor to buy a little closer to the bottom when a favorite stock retreats in price.

[NilesMike]

MACD is as illusionary as RSI, unfortunately.

http://etfhq.com/blog/2013/02/26/macd-test-results/

[Be Here Now]

I do not use RSI for sell decisions. Not being a trader, my infrequent sell decisions are based primarily on valuation, and then some combination of distances from one or more moving averages.

I rarely use RSI for buy decisions because the extreme oversold conditions that RSI can signal are so rare. Readings below 20 almost never happen in the stocks I follow, but when they do, I check my pulse and then look over my buy candidates list and my cash balance to see if I want to buy. For example, at the start of this month, both MO and PM had readings under 20, but I bought MO because it met my valuation criteria and I wanted more, but I did not buy PM because although it met my valuation criteria, I did not want more. Another example is LINE, which in early July of last yet had its RSI go below 20, but I did not buy because although it met my valuation criteria, I did not want more.

And yes, RSI like all technical indicators is right except when it is wrong. That is why I never use it in isolation.