RRG - RS and Mom seems to go in opposite direction of relative price

kiran

In the chart below, The RRG RS-Ratio and RS-Momentum lines are trending down in the last few days for AAPL. Why is that?
The relative price (AAPL:SPY) has been consistently trending up.

This seems deceptive - how is the RS-Ratio and RS-Momentum indicators calculated? I understand they're lagging indicators of relative price, but i don't see relative price leading these indicators downward.

markd

I'll take a shot at this until JdK shows up. I'm not a regular user of this tool (you know what they say about old dogs...), so take it for what its worth.

The chart school article does not show the calculations for JdK Ratio and JdK-Mom, so maybe they are proprietary.

It appears to me that JdK RS-Ratio is itself a measure of momentum, that is, incorporates time, whereas Price Performance does not. Price Performance uses the difference in price between two dates for two securities and plots each day's results. I'm guessing here, but it seems that JdK RS-ratio essentially computes a rate of change for price performance. If Price Performance continues to move up, but more slowly, JdK RS-Ratio will turn down.

JdK RS-Mom seems to be a rate of change for JdK-Ratio, so it turns down when JdK-Ratio advances more slowly.

Julius_RRG

Hi Kiran and Markd (thanks for jumping in),

The JdK RS-Ratio is a measure of relative trend and in the chart above it has picked up a new relative uptrend for AAPL against SPY when it crossed above 100. That situation has not changed, RS-Ratio is still > 100 indicating that the relative uptrend of AAPL against SPY is still intact.

(Note: be careful to mix benchmarks! In this case, it is SPY and $SPX used together which is not a big problem but in other cases, it may give weird results)


What DID change, however, is the acceleration (momentum) of that uptrend, you can fairly easily eye-ball it in the AAPL:SPY (I call that raw-RS) line in the lower pane.

During most of January, RS-Ratio remains flat but above 100 with RS-Momentum around 100. This is a stable uptrend. Then AAPL jumps which leads to a rise in RS-Momentum (both lines are based on trend-following principles so there is always a bit of lag) dragging RS-Ratio higher.

Then from mid-Feb the acceleration of the relative uptrend starts to fade which causes

• RS-Momentum to drop back and
  
• RS-Ratio to decline (because the trend it is measuring just got less strong)
  

and that is what you see in the chart

The interaction between JdK RS-Ratio and JdK RS-Momentum causes the rotational patterns on RRG-charts and positions securities not only against the benchmark for the universe but also against all other securities in that universe, providing you with the big picture.

In this particular case, the relative strength for AAPL (vs SPY) still looks good. The RS-Ratio is still well above 100 and the RS-Momentum line has corrected but is already leveling off. This puts AAPL in the weakening quadrant but with enough room to maneuver and turn back up towards the leading quadrant and continue its relative uptrend against SPY.

Hope this helps, Julius

kiran

This helps .. however, noticed that the downward turn in RS-Ratio and RS-Momentum around Feb 13th is very sudden, even though the Relative Price is smoothly decaying its upward movement, which seems strange.

Also, i'm trying to assess the approximate lag (in days/weeks) in RS-Ratio and RS-Mom to the relative price trend as a thumbrule. I realize the calculation is proprietary, but would be good to have a thumbrule to make some conclusions about the trend from the visualization. From the chart above, the lag on the Feb 6th jump-up seems quite small (1-2 days), however the sudden downturn in RS-R/RS-M on Feb 13th cant be explained from the relative price curve.

thanks
Kiran

kiran

Any thoughts/response to my question above?

Julius_RRG

Hi Kiran,

The RRG-Lines do not use the classical way of measuring relative performance by simply measuring the performance over a fixed period of time and comparing that to a benchmark or another security. There are multiple factors, and the interaction between them, involved to pick up the trends but sudden jumps in price/relative strength always cause jumpy behavior in derived indicators.

As a rule of thumb for the duration of trends that the RRG-Lines or better the JdK RS-Ratio line picks up, I have always used three months or twelve weeks when looking at weekly charts. Translating that to daily charts (5 days in one week) comes down to roughly 2.5 weeks.

Be aware that this is more of a minimum expected duration for a trend, many last much much longer

Julius