The new trade widget has the same functionality as the one you are familiar with from Rounds 1-5, with an added dimension.
You will have a better sense of the point cost/gain of a trade as you adjust the slider. The balloon value shows the adjusted probability of replication; the red curves show how many points you will tie up in the trade; the green bands show potential gains – again in market points.
As the probability changes, the cost changes in a non-linear way. Larger moves cost more — they always have, but now it is easier to see.
The boundaries (here 3% and 96%) are determined by your remaining points. The more trades you make, the narrower those will become. If you have shares in this claim, the boundaries will reflect your ability to sell those shares.
An example
Say you want to make a trade that reduces the current value of 40% to 20% (as in the image above) in a claim in which you have no previous stake.
The Basics
The gain/loss shows what happens to your trading points on each outcome, and considering only this trade.
You are reducing the chance of “Yes”, so you’ll lose points if the outcome is “Yes”. You have to pay those now to cover the possible loss. So if you make this trade, you will immediately be out 100 points.
You are increasing the chance of “No”, so you’ll gain points if the outcome is “No”. You’re not paid in points, but this matters in two ways:
- You are buying shares in “No”. (In this case, 142 shares: 100 to cover the cost you just paid, and 42 to pay the gain.)
- You can later sell those shares to “cash in” and recover points by moving the probability towards Yes. (See below.)
Recovering Points
If you make the trade and immediately put it back to 40%, that’s like doing nothing – the two trades cancel out. So you would recover exactly 100 points and be back to an Available Balance of 400.
If the market moves (nearly) all the way to No, you could sell shares to gain more than 100 points. If you sell them all, you would get (nearly) 142 points.
If the market moves towards Yes, you could sell to cut your losses. In this case you could recover only a portion of the 100 points to spent.
At the moment, you have to use one of the detailed views to see this. The text view is clearest. Here is what it shows for immediately reversing the trade – note the “Effect” section:
Cost Breakdown: ¿Trade vs Share?
You may have noticed the cost curve is divided into “Trade” and “Share” costs. You can ignore that if you like – the total cost after all is just what you get from the LMSR equation: gain = 100 * log2 (new probability / old probability). But some shares-oriented traders find it helpful to imagine you bought your shares at the price shown.
When you move the slider from 40% to 20%, you buy 142 “No” shares. The price of “No” was 0.60. So the ideal “shares” cost is 142 x 0.60 ≈ 85. The remaining 15 is the “trade” cost – you can think of it as the cost to move the price of No from 0.6 to 0.8 (or the price of Yes from 0.4 to 0.2).
Let’s look at the Loss Breakdown in Table View:
The total cost is 100 points. You pay that now, so your Available balance will change from 400 (before the trade) to 300 (after).
Share costs are determined by the starting price. The cost of No was 0.60. So your 142 shares would cost 85 points (if the price stayed constant).
Trade costs is the remainder: it is not a fee or friction, because you recover it completely if you undo the trade. It’s the cost of changing the price. Strictly speaking, the price is only “good” for an infinitesimal purchase.
Each share you buy changes the price, becoming progressively more expensive towards the extremes. “Trade costs” wraps that all into a single number – the cost of changing the probability/price.
