Accounting precision and effort

Tuesday morning report: 178.8B
Wednesday morning report: 177.9B

Ouch, in an average day I got 1.2B recently. So -0.9 is actually 2.1B loss. Nice, that needs some serious failure, like a buy order for 2x100M typed in with an extra zero. But no failure happened and I'm pretty satisfied with my Tuesday performance. What happened?

The cycle of an item is the following:

1. I have ISK. It's booked on its value, no doubts here
2. I set buy order, I lose the broker fee, the order is booked on its value, no doubts here
3. I receive the item, it's in my hangar, booked according to CCP algorithm
4. I set up a sell order, it's booked on its price, which can change (usually downwards) multiple times
5. I sell the item, have ISK, it's booked on its value, no doubts here

With an example:

1. I have 100.2M ISK, booked as 100.2M
2. I set buy order for 100M, 0.2M broker fee, booked as 100M
3. Item is in my hangar, CCP algorithm says it worth 90M, so it's booked as 90M
4. I set up a sell order, for 120M, it's booked on its price, 119.7M (0.3M broker fee)
5. I update the sell order, to 115M, it's booked on its price, 114.7M
6. I sell the item, have ISK, 114M (tax)

The endpoints are dead certain. The sell order part is OK-ish because I sell 25-30% of my wares every day, so my prices aren't much above equilibrium. The nasty part is the CCP algorithm. It is a very-long-time average of the whole galaxy. Since I mostly deal with implants that were butchered by the FW design disaster, their CCP value is often 75-85% of their Jita buy orders. So if I buy implants with buy orders for 10B, according to the algorithm I lose 2B when the orders fill up. The implants surely worth as much as Jita buy orders (-tax) as I could instantly sell them for it.

On Tuesday I increased my trading volume. I did not include new items, because that takes time, but I decided to have 10 of each, so I can sell more with the same amount of clicks. I got them, hence the drop in the total assets.

Of course I could abandon this clearly flawed algorithm and use 3-days average Jita middle price for every single item I have. But that would be awful lot of counting.

The question is, can I afford to use a clearly bad method of accounting? The answer is yes because I have high rotation. The time between the first and the last steps (the ones that are pure ISK and perfectly booked) is few days. In simpler words: it doesn't matter what value the implants are booked as they'll be sold in a few days anyway.

If I were doing long-time investments, this would be horribly wrong. Imagine to have 100B booked items with the real market price of 70B. Everything is fine and shiny until I do the liquidation and bang, lost 30B overnight. In my current trading method I buy and sell relatively the same amount every day, so the errors from the bad booking are canceled out.

There is one problem though: total value assessment. If I were to stop trading and move to other field, the final value (which is dead-certain ISK), could be seriously off from the current guess. That's bad. But I'm not planning this. I'm using the assessment for measuring growth and the errors between the ends of months cancel out.

---

Thursday morning report: 178.9B, 8.6 spent on main accounts, 7.1 spent on Logi/Carrier, 3.8 on Ragnarok, 5.5 on Rorqual, 3.4 on Nyx, 3.4 on Dread, 57.4 sent as gift)