Basics chart analysis 2/3

The donchian channel
The donchian channel was named after his inventor, the economist Richard Davoud Donchian, who first named such a channel in the article “High finance in Copper” in the “Financial Analyst Journal”.

At this point, I will just cite wikipedia, as the explanation is quite good:

“[The donchian channel] is formed by taking the highest high and the lowest low of the last n periods. The area between the high and the low is the channel for the period chosen. […] If a [good] trades above its highest n periods high, then a long is established. If it trades below its lowest n periods low, then a short is established.”

(For everyone who is not accustomed to the terminology: you can read “going short” as “selling” and “long” for “buying”, then it should make a bit more sense.)

Transfered to EVE, this means if we connect the highest prices and the lowest prices of the last 100 days, we will create an area which is called the “donchian channel”. Easy, right?

Right now, the donchian channel is shown as a grey area – I highlighted it a bit on the right side… dont judge my paint-skills, please

Now we have the donchian channel and the question is, how we can use it in our decision-process in regard on how, when and what we trade.

First it is important to say, that the donchian channel is an indicator for the stability of the price of a good. In EVE as well as in reality, this means that a “small” donchian channel is a sign of a more stable good, while a “wide” donchian channel means that the price of the good is way more volatile, the prices have a higher fluctuating. If you e.g. are more into stationtrading, a wider donchian channel is a bit more primissing, as there may be a higher spread between buying and selling, meaning that your profit from pure arbitrage-trading is better. But for this, it would be better to use the “minimum/maximum”, as the donchian channel in EVE is not real-time but delayed by 5 days. If it reaches a new high/low, it uses the actual number, otherwise it will only change if 5 days pass.

In the case of EVE, I need to say that the analysis of the donchian channel comes to a kind of an end. While it is still in use in real life (I wont dive into the discussion if this is adviseable or not) for the analysis of trends or for the exact point in time, when to leave or enter the market. Basically, it is used like this:

  • buy, if the moving average (5 or 20 day is a case-to-case – decision) reaches the top of the donchian channel and
  • sell, if the graph reaches the bottom

This sounds a bit strange at first: buying when the graph reaches an high and selling if it reaches a low. But we need to take into account that the donchian channel is there to measure trends. “Trend” means that the price is rising of falling over a period of time and we use this information to sell our product in the future at a higher price than we bought it for. It is a classical investment, which means buying –> holding –> selling and not an arbitrage activity, which utilize more the difference of price at the same time, like we do in the classical stationtrading.

But EVE is still a game and the margins as well as the volatility of the goods we trade with are rarely shown in real markets (highly speculative exceptions aside) in the amount and time. Every patch, every change in the meta influences a good and can either let it skyrocket or fall down like a stone. Of course, this can also happen in reality, but here the conditions are fundamentally different. A market in real life can go down as a whole (last seen in the financial crisis 2007/2008), as there are many complex interdependencies and connections (state politics, securities, mentality, regulations on a local as well as on a international level etc.) between all participants on the market. In EVE, there is no such event as all markets going south as a whole, but only single products or at most a couple of groups of products, as the connections between these products are more known, less influenced by these interdependencies and most of all: known to everybody. As long as you have no stockpile, ridiculus amounts of a good or a monopoly (like the CFC did with technetium some years ago), it is hard for you to influence something in the long run. One way would be of course manipulation of the crowd via e.g. faking a news from CCP on twitter, but this falls more on the section of scamming and does not require analysis of financial tools like the ones we talk about here.

Anyway, I start to digress. Investment in EVE is also working different from investments in RL in the first place. While in RL you rely on instruments like the donchian channel and in some cases advisory from e.g. scientists who are working in this field, this role is mainly distributed to patches and their conclusions. One example for this is Salvage. Compared to the price of last year, the prices of salvage went bonkers and will most likely rise up even more in the future. The reason is simply logic. Because CCP has announced to remove the POS-system a long time ago. Producing then is only possible on Citadels and Engineering Complexes. But Citadels and Complexes need Rigs, just like a ship. And Rigs need a huge amount of salvage – the bigger they are, the more salvage they need. There is no need for a donchian channel to know that the prices for salvage will rise up a bit more during this year and even more when the point in time when POSes will be obsolete is finaly public. Same goes for Meta-4-Items and any POS-Module which is unprofitable to produce right now, as the payment CCP gives out for these modules might be oriented on the resources needed to build them.

Basics chart analysis 1/3

As I had a very positive feedback on the german version of this post, I thought about translating, updating and expanding the part on chart analysis in EVE. The first two parts I will simply elaborate the informations that are available ingame, their usage and some history. In the third part, you will find some techniques I use / used with third party – tools and excel.


EVE is also called “Excel in space”. The reason for that is obvious, because you encounter tables in every single aspect in the game, be it mining, trading, production, PvE or PvP. A newbro rarely gets a grasp on these tables or what they might tell him – it needs a bit of learning and thinking about what he has on the screen, after which he can use this information.

Getting informations from the graph ingame itself is not always easy, but concluding the correct informations is even harder and most the time a bit of an issue of instinct, which comes with time and experience. At first some words on the graphs itself: it shows the market data of one year of the most famous item of EVE: the PLEX or the “pilot licence” for capsuleers. Recently, this item suffered from a spike upwards, as… well, there are many theories and countertheories out there – it is up to you, which one you believe.


This graph however contains several informations: the traded volume per day, the moving average in 5 & 20 days, the median day price and the donchian channel.

1.) Volume traded
The amount of units traded every day is shown at the bottom of the diagram. This information alone is of little use, as it alone can only be an indicator of a stronger or weaker demand and therefore rising or sinking prices. For a more distinguished picture, we need to take into account one more information which is not shown in the diagram, but in the table alone: the amount of orders under the “table”-view of the diagram.

You can – in theory – get the amount of orders (which is a bit misleading, as it is more the “amount of transactions done today” and not “amount of orders currently on the market”) and calculate an “average amount of items sold/bought per order”. I said “in theory”, because this number is only of little use, as long as you do not get data dumps every hour or so. If you do this, it might be possible to locate regular rush-hours for this specific product.

2.) Moving average
The moving average is nothing more than the information about the unweighted mean over the last 5 (blue) and 20 days (orange). This means, that the price of all orders of the last 5 or 20 days will be added up and then divided by 5 or 20 – simple and easy, just as we all (hopefully) learned in school. This means, that the orange line with the mean of the 20 days will react way slower to changes in price than its counterpart. The picture shows this pretty good: the blue line always reacts faster than the orange one. If both of these graphs rise up, it can mean 2 things: that the demand is higher than usual or that the supply is lower than usual. Which one of these situations is correct does not really matter in EVE, but the consequence is the same. Either you start to produce the good now, speculate the on that the price is rising even further or do whatever you like to do with this information.

3.) Median day price
The median day price is a simple information, but other than the concept of the moving average, the foundation of the median is not always taught in every school. And also we need to keep in mind that not everyone in EVE attended school recently or is working with mathematical formulas in his everyday life. In our diagram, the median day price is shown as small orange dots which are not connected with each other. 

Anyways, the median day price is easily explained with a simple example. So let´s imagine that I have 7 Kestrels (a small caldari frigate) and sell one of them each day of the week at the following prices:
Monday: 1 million
Tuesday: 2 million
Wednesday: 3 million
Thursday: 2.5 million
Friday: 1.5 million
Saturday: 1 million
Sunday: 170 million (as a customer accidently forgot the komma)

While the average take the sum of all these sales divided by the amount of the results, the median is somewhat of the “middle” of all sorted transactions. The average price would be totally messed up by the one las tlucky transaction, meaning that we sold our Kestrel for an average price of (1+2+3+2,5+1,5+1+170)/7 = 25,86 million ISK. Obviously, this information is not useable.

You will see that the median on the other hand is quite robust against such exceptions. To see this, we need to sort the amount of data, beginning with the smallest number: 1, 1, 1.5, 2, 2.5, 3, 170. The median is the exact center of this numerical sequence: 1, 1, 1.5, 2, 2.5, 3.0, 17. So this means that the median of the same data is far more realistic than the average of about 25 million. Thats the way the median price works – but not on a weekly basis but on a daily one. It is easy to say that the median day price is one of the most hardest one to significantly manipulate as a single player.

The last two informations “donchian channel” and “min/max” will be the toppics of the next entry.
Fly dangerous o7.