To remedy this problem, the simExchange began the use of an experimental new IPO process to determine the IPO price. We evaluated many IPO processes such as Dutch auctions and Parimutuel Auctions but decided on a Double Call Auction to keep the process simple. A double call auction is how trading on the simExchange has always functioned. The only difference from standard trading is that the automated market maker is turned off.

Price discovery works through traders submitting buy orders at the maximum price they are willing to buy at and submitting sell orders at the minimum price they are willing to sell at. Essentially, traders are posting a range of how well they think the game will sell—a maximum and a minimum.

For a single trader, this range may be very large. They may think a game will sell a minimum of 400,000 copies (40 DKP) and a maximum of 600,000 copies (60 DKP). This would equate to a buy order at 40 DKP and a sell order at 60 DKP. With this player’s orders alone, the bid-ask spread is a very large 20 DKP.

However, a second trader may believe the game will sell at least 500,000 copies (50 DKP) and at most 650,000 copies (65 DKP). The trader’s orders represents a 15 DKP bid-ask spread, but together with the other trader, the best bid is now 50 DKP and the best ask is now 60 DKP. The bid-ask spread is now 10 DKP.

A third trader believes the game may only sell 300,000 copies (30 DKP) with a max of 550,000 copies (55 DKP). The bid-ask spread is now 5 DKP. The order book would look like this:

As more traders submit orders, the bid-ask spread will tighten and converge on a market price. Individually, no one knows what the IPO price should be, but together, traders can narrow the range down substantially. This price range should be the best guess of a fair IPO price as this is where the buyers and sellers meet (the optimists and pessimists for the game’s potential sales). In this IPO process, trades only occur when traders are willing to buy and sell at the same price. Once again, this is no different from normal trading except there is no automated market maker submitting orders.

However, this process has proven to be difficult to understand for many members of the simExchange, especially those who use the Basic Trading mode, which is limited to placing market orders (orders that immediately take the best available price). If a trader is not used to looking at what is the current selling price, he may be in for a surprise when his buy order fills.

Originally we had considered reserving the IPO process only to those using the Advanced Trading interface so that traders are forced to identify the matching price they would accept. However, we thought traders using the Basic Mode should still be allowed to participate if they see a price they think is good for buying or selling. Of course, this assumes the trader using the Basic Trading mode is paying attention to the current bid and ask prices. However, many traders using the Basic Trading mode ended up with prices they were not happy with.

Last night, we have heard a deep debate regarding the IPO process. Our goal is to make accurate forecasts in an entertaining and easy to play process. It appears the experimental IPO process has failed to accomplish those goals for many traders.

Based on user suggestions, we will combine the experimental process with elements of the original IPO process into a 2-step IPO process. Once a stock or future is listed on the simExchange, there will be a stage to determine the IPO price. This will be a three-day period in which players place orders in a double call auction. To participate in the double call auction, the player must specify the maximum price he will buy at or the minimum price he will sell at. After the three-day period, an IPO price will be calculated. At this point, the second stage will occur with the stock or future available for purchase or shorting at the IPO price all day by all traders. Following this day, regular trading with automated market maker will commence.

This article was cross-posted from Discovering an IPO price dated July 26 on The simExchange Official Blog.

After observing a few months of trading on the simExchange, the NPC market makers have been tweaked to provide more consistent liquidity across stocks and smaller price differences between their orders. The effect of these changes should result in more activity in higher priced stocks, such as console stocks. Percent price changes will not be so erratic. Prices should no longer jump between down 3% for the day to up 2% of the day. Prices should change more frequently but in smaller magnitude at a time. [...]

Ah.

Previous: The Sim Exchange: Stocks vs. Stock Derivatives (Futures & Options) + Keith Jacks Gamble: simExchange is somewhat OK, but will remained confined in play-money land. + Brian Shiau: The Sim Exchange Works Fine, Thanks. + Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest + The structure of simExchange game stocks + An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

Brian Shiau:

[...] I also believe gamers think of game sales in terms of lifetime sales. People remember that Halo 2 sold over 7 million copies. They don’t remember how many copies it sold between months 2 and 4 of release. They are thinking Halo 3 as an 8 million copy seller, not a 3 million copy seller between weeks 3-6 of release.

[...] Once a game nears the end of the lifespan and the total global sales has been predicted to a certain degree of accuracy, there is no longer a need to continue trading it. When the lifetime sales of a game is fully priced into a stock, players will cash out and turn their attention to new games in the same way growth investors cash out of mature companies by selling their shares.

[...] I believed it is more clarifying for players to understand that they can buy and sell at certain prices and quantities because another player is offering that to them. [...]

#2. THE SIM EXCHANGE: Stocks vs. Stock Derivatives (Futures & Options)

Brian Shiau:

[...] This is part of the reason stocks on the simExchange are called “stocks.” When we launch a derivatives market for options and futures, we will call them “options” and “futures.” Probably a rumor slightly overshadowed by the black Xbox 360 Elite, but yes a derivatives market on the simExchange is coming. Remember: stocks do not expire and cash out on a regularly scheduled basis. Futures and option contracts do.

[...] Short selling is borrowing stock and selling it, hoping to buy it back at a cheaper price later and pocketing the difference. [...] Short selling is different from buying a put option or writing a call option—these concepts should not be confused and we would not be doing players any favors by mixing them up. [...]

Previous: Keith Jacks Gamble: simExchange is somewhat OK, but will remained confined in play-money land. + Brian Shiau: The Sim Exchange Works Fine, Thanks. + Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest + The structure of simExchange game stocks + An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

2. Ever wonder what the rest of the world really thinks about the games you already pre-ordered five months ago?
The simExchange lets you observe an aggregated forecast for how many copies a game will sell globally. Experts explain that market-driven systems can aggregate information from thousands of people, resulting in a better forecast than from any single analyst.
3. Want to find the games that are really worth your time to read more about?
The simExchange enables you to find new video games through the “wisdom of crowds.” By aggregating the opinions of thousands of gamers, the simExchange saves you time in identifying the most anticipated games simply by checking out the stock price or trading activity for a game.

Previous: Keith Jacks Gamble: simExchange is somewhat OK, but will remained confined in play-money land. + Brian Shiau: The Sim Exchange Works Fine, Thanks. + Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest + The structure of simExchange game stocks + An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

Thanks for the response. It’s interesting to see examples of product news stories and how your markets responded. These examples suggest that your game share prices are connected with sales. I’m not surprised and Keynes wouldn’t be either. His beauty contest view explains exactly why prices on the simExchange are connected to sales despite the fact that game shares have no intrinsic connection to sales (no dividends based on sales, nor the possibility to liquidate based on actual sales). The traders’ comments you mentioned confirm that traders have picked up on this point and are buying and selling in anticipation of other traders’ actions. Certainly, a lot of trading on Wall Street works the same way.

My point that game shares have no intrinsic value, unlike Wall Street shares, has two implications. First, it’s one reason that prices on the simExchange may deviate more from actual sales than prices on Wall Street exchanges deviate from actual value. Importantly, this statement doesn’t say that simExchanges prices will deviate more, nor does it say that any deviation will be large. Further, your simExchange has at least one advantage for keeping prices near sales that Wall Street does not have: your market makers have infinite resources to keep prices at reasonable levels. Second, although irrelevant since the simExchange uses play money, the fact that game shares have no intrinsic value prevents the simExchange from ever working with real money.

Previous: Brian Shiau: The Sim Exchange Works Fine, Thanks.

Previous: Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest – by Keith Jacks Gamble

Previous: The structure of simExchange game stocks

Previous: An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

It is a valid criticism, since fundamental analysis is essentially impossible until sales data by NPD Funworld is released (Shiau has some nice empirical results relating to prices ex-ante and ex-post sales figures, and the resulting market corrections…scroll down until you see the charts). However, this is largely a short-term problem and I suspect will disappear in subsequent weeks. With more companies offering demos of games as downloadable content, market participants have access to better information. This reduces the amount of “noise” created by previews and reviews by the gaming media, since 1) the quality of the reviewer varies and 2) most of their work covers unfinished builds of games, which may be materially different from what is released to the market.

Previous: Brian Shiau: The Sim Exchange Works Fine, Thanks.

Previous: Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest – by Keith Jacks Gamble

Previous: The structure of simExchange game stocks

Previous: An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

[...] Many players realized early on that share prices on the simExchange are completely based on what other people are willing to pay: [...]

[...] Empirical evidence of trading on the simExchange has demonstrated at least some [market] efficiency. [...]

Players who did not trade in the direction implied by the news lost DKP while those who did gained DKP. As predicted by market theory, eventually, these players learn to trade with sales figures or they run out of money. However, like the real stock market, the simExchange is not immune to bubbles. If people become irrationally exuberant for a game, the stock may be priced very high. This is all part of simulating a stock market.

Additionally, as in any game, the admins can wield the awesome power of Non-Player Characters (NPC), in this case market makers with infinite money, that can encourage players back to rational trading.

Previous: Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest – by Keith Jacks Gamble

Previous: The structure of simExchange game stocks

Previous: An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

Er, it sure sounds like they don’t enforce any connection at all at any date between the game mentioned by the asset and anything related to sales of that game. If there are not enough traders of good will to enforce such a connection, then with learning the connection will probably be lost.

Previous: The structure of simExchange game stocks

Previous: An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

Contracts on most prediction markets are often binary contracts that pay depending on whether the event described by the contract occurs or does not occur. This structure is often referred as a binary option [1]. However, a prediction market is not restricted in solving yes or no questions. Contracts can be created to pay a scaling amount so that a prediction market can attempt to ascertain a quantity, such as how much a movie will gross in box office receipts.

Prediction markets have used contracts similar to futures to answer such questions. These contracts have some value that corresponds to a prediction and expire at a certain point, such as four weekends after a movie is released. At expiration, the contract holder cashes out the contract at the spot price (or the sum of box office receipts after four weekends) [2]. However, the problem with video games is that a game can continue to sell for years to come so any arbitrary expiration is not indicative of how well a game will sell. In creating a prediction market for video games, the simExchange required a structure that would accommodate the nature of the video game industry.

There are other quirks to the video game business. One particular problem that has been the ire of many analysts is the lack of comprehensive sales data [3]. Unlike Hollywood movies, video games do not have official sales figures every weekend. Instead, the industry relies on point-of-sale studies, surveys, and intelligent extrapolations from companies like NPD to get an estimate of how many copies a game has sold. The number of copies a game has sold will vary from source to source, although NPD is considered the standard by many in North America as it is the most comprehensive for North American sales.

Given these two problems, video games can continue to sell for years and the lack of data with absolute truth, the simExchange could not easily adopt the structure of most prediction markets already in existence. Instead, it sought a time-tested structure that has been used to answer a similarly mirky question: how much is a company that may last for decades really worth?

No one knows with absolute certainty how much a company is actually worth. That is one reason why the stock market exists for people to trade shares of a company. The stock market aggregates the information of all the traders to hopefully ascertain an accurate valuation for the company (this concept is known as the Efficient Market Hypothesis [4]). Due to the similarity of the issues, stocks on the simExchange function very similarly.

A stock’s price on the simExchange corresponds to the lifetime worldwide sales of a game, in which 1 DKP corresponds to 10,000 copies sold. These stock prices will climb or fall with monthly sales reports, just like a company’s stock price will climb or fall with quarterly earnings reports. A stock on the simExchange will also increase or decrease as a result of news on the product, just as a company’s stock will increase or decrease as a result of news on their products. If people believe a stock is underpriced given the data, people will bid it up and vice versa [5]. There is no automated function by the New York Stock Exchange to cash out a stock and pay shareholders a lump sum of cash depending on how the quarterly earnings for the company fared.

Eventually, a game will stop selling, just like eventually a company will stop growing. In this case the stock price will merely stagnate. Investors of game stocks can cash out just like they would with company stocks by selling their shares (or covering if they are short the stocks). The simExchange market makers will supply the liquidity to close those positions.

Due to this structure, in an efficient state where a diverse pool of traders are participating in the simExchange, game stock prices should become a strong predictor of the lifetime worldwide sales of video game titles [5].

References:
[1] Wolfers, Justin & Zitzewitz, Eric. “Prediction Markets in Theory and Practice.” March 2006. (PDF)
[2] Hollywood Stock Exchange Frequently Asked Questions.
[3] Electronic Gaming Business, October 6, 2004.
[4] Shleifer, Andrei. Inefficient Markets: An Introduction to Behavioral Finance. New York: Oxford University Press, Inc. 2000.
[5] Chen, Kay-Yut & Plott, Charles R. “Information Aggregation Mechanisms: Concept, Design, and Implementation for a Sales Forecasting Problem.” Hewlett Packard Laboratories and California Institute of Technology. March 2002.

Originally published on the Sim Exchange website. Republished on Midas Oracle .ORG with Brian Shiau’s permission. ((( Appreciated. )))

Related: Keith Jacks Gamble: simExchange is somewhat OK, but will remained confined in play-money land. + Brian Shiau: The Sim Exchange Works Fine, Thanks. + Robin Hanson on the Sim Exchage + simExchange a Keynesian Beauty Contest + The structure of simExchange game stocks + An invitation to join the simExchange beta + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + Sim Exchange – How to earn additional money? + The Sim Exchange: Basic Trading vs. Advanced Trading + BetFair, Sim Exchange = Vertical Prediction Exchanges, First

Previous: BetFair, Sim Exchange = Vertical Prediction Exchanges, First + The Sim Exchange: Basic Trading vs. Advanced Trading + Since November 9, 2006, the Sim Exchange has attracted over 2,400 registered players. + An invitation to join the simExchange beta