Deceptive Divergence

I’ve seen a lot of folks on crypto twitter posting the following chart and pointing to the divergence, stating it’s a reason to get long:

First, as I always say, divergence on its own is never reason to get long or short. Divergence is a supporting criteria, not a primary one.

But more importantly, this isn’t valid divergence. The scandal!

Most new traders don’t understand it. Lower low on price and higher high on RSI, right?! Nope, sorry… not that easy.

Some more experienced folks will tell you that it’s not valid because it’s “too far apart”. And while this is kind of true, it’s not the whole story. Folks saying this are either saying, “I don’t really know why it’s not valid, but I want to sound like I know what I’m talking about”. Or, “It’s too nuanced a conversation for me to have right now”.

Well, here’s the nuanced version for those that care to learn.

For starters, let’s zoom in on the price action in question:

You still see our “divergence” lines, but I’ve also labeled the highs and lows with the corresponding letter on RSI.

When we break it down like this, it’s easy to see what’s going on. After the drop, bitcoin put in two green candle and then a red one. This marked our first high, point ‘A’. This was later followed by a low (point ‘B’) a higher high (‘C’), a higher low than our low at point ‘B’ (point ‘D’), a higher high than point ‘C’ (point ‘E’), and finally a lower low once again (point ‘F’).

You could probably draw a rising wedge in this series, a bear flag… whatever. The point is there is a clear leg of higher highs and higher lows here. It’s an uptrend.

Until point ‘G’ where we failed to make a higher high but we made a lower low. Point ‘G’ must be compared to point ‘F’ because ‘F’ was the last low we put in. In no world would you compare ‘G’ to ‘D’ or ‘B’. It just doesn’t work that way.

Price has started a new leg, and we see that clearly as we put in a lower high (‘H’ is lower than ‘E’), a lower low (‘I’) and what may be another lower high (‘J’).

What folks are incorrectly doing is comparing the low at point ‘I’ with the low just before point ‘A’. But the price action starting just before point ‘A’ is an uptrend and we’re now in a downtrend. Again, you can’t correctly compare ‘I’ with the low before ‘A’.

You need to be comparing ‘I’ with ‘G’, and if we do that, we see that RSI indeed put in a lower low as price put in a lower low. This is not a valid case of divergence, and that’s what I meant when I tweeted this:

You can clearly see this misstep when you look at the 15-minute chart. Would you say this is divergence?

Clearly not!

But things get even more interesting. At this point we can definitely say that the 1-hour divergence is not valid. When you tell someone that, the next thing that’s likely to come out of their mouth is, “Well just look at the 4-hour chart then!”.

So let’s look:

Yes, at first glance this looks like some pretty solid divergence. The problem for me is that I generally see divergence develop on lower timeframes and sort of “trickle up” to the higher ones. As we just discussed, there’s no divergence on the 1-hour chart or the 15-minute chart. So the “trickle up” theory isn’t really holding up here.

Hmmm…

What about the daily?

Price made a lower low on the daily and so did RSI.

So the 4-hour chart is the only one showing “divergence”.

Confluence is a power tool. When multiple support and resistance areas meet in the same spot, it makes that area all the more powerful. The same holds true for trends. If the 15-minute, 1-hour, 4-hour, and daily charts are all in downtrends, that’s a damn powerful trend.

But if you can only identify a downtrend in the 15-minute chart while the 1-hour is neutral, the 4-hour is bullish, and the daily is bullish… you better think twice about putting that short position on based on the 15-minute chart. At a minimum recognize that the odds are less in your favor than if you had multiple timeframes aligning to confirm the trend.

The same is true for divergence. The fact that we see divergence on the 4-hour chart and only the 4-hour chart is a weak indicator at best. Again, I wouldn’t take a trade based solely on divergence, and now we’re looking at a weak, unsupported case of divergence? No thanks.

The more likely scenario is this is a false positive. And this is why I stress to folks that they should always be charting and looking at multiple timeframes, preferably the 15-minute, 1-hour, 4-hour, daily, and weekly (throw the 5-minute in there if you really want to zoom in).

In short, this divergence on the 4-hour chart isn’t bothering me too much. I’m still holding my shorts and waiting for a better indication that it’s time to get out.

Geppy

I've been involved in the forex markets for over a decade, initially starting as an FX trader at Allston Trading in Chicago. Eventually I went on and founded my own (non trading related) company. I spend my days working from home and trading forex, equity, and crypto markets.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.