Relative Strength misunderstanding

[ryanmackk]

Hey guys, this is funny and slightly embarassing but I don’t fully understand relative strength. I need more than the wikipedia/investopedia explanation. I understand what the RSI is: a comparison of a stock’s performance relative to the market , so if RSI is 70, the stock has performed better than 70% of stocks in the sample market.
But I don’t understand what Relative Strength is… lol
1. If a stock in increasing in price and so is the market, is it relatively strong?
2. If a stock goes up while market dips, is it strong?
3. If a stock goes down and the market dips, is it strong?
I thought relative strength was just correlation analysis, like for instance -1 would be strong and 0 would mean weak….. but apparently, as I’ve heard recently, if the correlation is 1, and the stock is falling WITH the market… it’s weak… and if the stock is rising and the market is as well with say correlation 1, it’s strong. So it’s not correlation analysis then…………I’m so confused

 

[S2000magician]

If you look at the definition of RSI:
RSI = 100 – 100 / (1 + RS)
where:
RS = average of n days’ up closes / average of n days’ down closes
you’ll see that RSI has nothing to do with the market; it’s simply a measure of a stock’s performance against itself.
Over the last, say, 30 days, has it been up more than it’s been down? Has it been up a lot more than it’s been down? Or has it been down a lot more than it’s been up? That’s what RSI is trying to measure.
No correlation. No market. Just GOOG vs. GOOG.

 

[ryanmackk]

ic……. This adds to the confusion for sure….. So wiki/investopedia are wrong in their definitions?

 

[S2000magician]

I got that definition straight from Investopedia.
Let’s see what Wikipedia has to say:
Same thing (though they give more detail on how the averages are computed: exponential moving average rather than simple moving average).
So, no: their definitions are correct.
Try this: go to Yahoo! Finance and look up … well, I used GOOG above, so let’s stick with GOOG. Download the prices into Excel (there’s a button at the bottom of the page: Download to Spreadsheet). Let’s use, say, 30 days, and to avoid problems with changing data, let’s go from 14 Aug to 25 Sep. In your spreadsheet, make a column for the daily price change, and compute today’s close minus yesterday’s close. Use the “Adjusted Close” column.
Here’s a check: I got -$9.61 for 25 Sep, and -$11.44 for 14 Aug. I got 16 down days and 14 up days.
Make an “Up” column, and put in only the positive price changes, put in zero for the others. (You can do this with an IF() function.) Make a “Down” column, and put in only the negative price changes (but change their signs to positive).
Here’s a check: for 25 Sep you should have $0.00 in the Up column and $9.61 in the down column.
Average each of the columns. I got $3.18 for the average of the Ups and $3.32 for the average of the Downs.
Throw these into the formula:
RS = Avg(Ups) / Avg(Downs) = 3.18 / 3.32 = 0.9596
RSI = 100 – 100 / (1 + RS) = 100 – 100 / (1.9596) = 48.97.
GOOG is square in the middle: not overbought, not oversold. RSI says “Hold”.
(This, by the way, is a lot more than you need for the Level I CFA exam. All you need there is that RSI compares the up days and down days for GOOG, if it’s above about 70 then GOOG is overbought (sell signal), and if it’s below about 30 then GOOG is oversold (buy signal).)