The dangers of Renko and pitfalls of money management

Did you ever see these wonderful charts, that dropped some time ago to your spam along a cheap trading system ?

renko

Renko

 

These are japanese Renko charts, or price movement charts. That is, Renko charts just show the movement of the price, regardless of the time.

They have a defined base, or brick, in pips. When the price movement touches the height of a brick, the indicator draws a brick on the chart, either bullish or bearish. And so on.

And they indeed look fantastic. Long straight lines, easier signals, it all looks like a charm?

Where is the deception ?

What if you consider bricks to be price events? From your point of view they may be as well roulette streaks of black or red. And who wouldn’t want long streaks on the same event ? But is it really this way ?

The thing is, IT SHOULD BE. Given the definition of the Renko, it is all pretty sound. A brick is drawn when the price makes the movement of a brick.

BUT THAT IS NOT THE CASE. Simply, because most platforms don’t have the technical capabilities to use historical tick data in indicators.

Therefore , a Renko chart is correct only if it is drawn on a runtime basis, tick by tick. Historical Renko is deceptive.

How big is the deception ?

Well, quite everything that you see on the screen is deceptive. Because Renko indicators usually take the open-close movement.

Say a bar has 100 pips movement up from open to close, with a 20 pip brick size. Historical Renko would display that bar as 5 bullish bricks.

But what if the 100 pips movement was on an 150 pip bar, where just the body is 100 pips in size? Where are the intrabar movements ?

Renko simply removes all the intrabar action. So, the final result will be 5 bricks up, but until they are like that they can go 2 down and 6 up and so on. Renko will actually redraw not one, but all the bricks that the current bar from the base chart is composed of.

Therefore, what you see on the screen is completely fake. Of course, good Renko indicators will make use of history at its maximum extent, that is, in most cases, 1 minute bars. But the tick interval is impossible to use on most of the platforms.

To make the matter worse, a Renko indicator may have a dual behavior, using historical approach for past data and tick-by-tick approach for current time. The trader wouldn’t suspect anything, as no brick will be redrawn.

“There is always a catch”

There is a tiny detail that escapes the eye when looking even at the most correct, tick-by-tick, Renko, and that is:

The price movement needed to produce a contrary brick is double the size of a regular brick.

As you can see, each brick is above/below the previous brick, never on the same level, as it would be the case for a simple direction reverse.

renko2

 

Of course, a simple direction switch from N pips up to N pips down can be drawn as in the regular Renko. However, that would break the meaning of the last Renko price level. If you draw it one level down, coming after one level up, then there are two levels difference between brick last prices, instead of just one, as it seems to be at first glance.

Sure, nobody likes a correctly displayed Renko chart with even movements. It would look as cluttered as normal charts. It would quite remove the magic about it, especially when range zig-zag movements would create a line of bricks of alternating colors.

Renko implications on money management

And here begins all the jazz. Opposed to coin flipping on roulette where coin flipping events are usable only within themselves, Renko delivers event extensibility, as it allows profit accumulation from the previous trades into the area of the current trade.

The uneven movement system also gives a powerful punch to the brick ending a streak.

A reverse brick is about: two levels for itself, one level for the previous brick, and flatten the last but one brick.

As in the example of “Renko uneveven movements”. We went long three times: first we won the first brick, then we went long again, we made the second brick, plus an additional level of profit from the first trade. Then we went long again. Market went down two levels. We lost two levels for the 3rd trade, one level for the 2nd trade, and flattened the first level. Output zero.

How would the trailing stops have worked?

Bear in mind that if they are higher, they can trigger before an opposite bar is created, thus affecting the meaning of Renko. Renko, applied correctly, with tick data, is somewhat a trailing stop strategy by itself.

Let’s suppose 3 movements up and one down.

After 3 movements up, running profits would be: 3+2+1 = 6 units. But only one reverse brick would spell doom: -2-1.

Therefore the question of playing the Renko, in the absence of sound charts that would allow taking correct decisions, would rather be where to cut the profits.

One might like playing the martingale with Renko, closing each trade when a brick is formed up. Due to its uneven movement feature, a Martingale streak will end of 2^(n-1)+1 profits instead of 1 unit.

For instance, in the first case, waiting for a red brick , would be as follows:

Step 1: bet 1, lose 1, total -1
Step 2: bet 2, lose 2, total -3
Step 3: bet 4, win 8, total 5

Certainly better payouts than regular martingale, but one has to be aware that:

  •  event is unevenly positioned, it requires a double movement to happen
  •  switching directions wouldn’t work: betting for a brick of the same color as previous won’t benefit the doubling profits feature.
  •  drawdowns will follow the same function as regular martingale.

Is Renko good or bad?

Thing is, we can’t know that. Each must do his homework. Get the data, mine the data, check for the edges while charting it. There are no shortcuts to success…

 

By | 2013-11-28T15:44:17+00:00 November 28th, 2013|Blog, Forex|2 Comments

2 Comments

  1. luka November 6, 2014 at 10:00 pm - Reply

    Excellent article! i have a question regarding renko charting. Anyway i searced whole internet in hope of finding “real renko chart” that can be applied in forex but i haven”t find one. I”m novice trader and find renkos quite useful, so my quoestion would be do you know anyone capable of programming such thing or is it even possible? Is there any info about someone succesfully developed it?

  2. Bogdan November 18, 2014 at 9:39 pm - Reply

    It would be quite difficult to program it, and probably it would be prone to crashes. From a regular trading platform that does not have tick data, it would need a large database – either as CSV files with ticks or a real database, accessible via SQL – and then fetching data, which would require DLL calls (for instance to MySQL) or going thru a huge text file (which would probably have to be parsed in the beginning).
    As we didn’t test this yet, it might be possible to be written inside JForex. But both ways – either having it as text file, or written in JForex, would require JForex tick data (as I can’t think about another source). This would be pretty painful, as managing the cache is extremely time consuming for the platform (and getting the data in the cache is required for both history download or the indicator working over the platform). It would be probably to simply give up on the indicator and just download the tick data, and then data-mine it externally, with the indicator working in a non-graphical
    manner.

Leave A Comment