7. Experimental Markets Flashcards
What are experimental markets?
Experiments in which:
-some experimental subjects are assigned roles of buyers and/or sellers
-there are goods that can be traded
Describe the vernon smith 1962 experiment
-subjects divided into buyers and sellers
-each buyer has a max price and each seller has a min price
-the min and max prices are private info of the individuals
-we watch the prices and quantities at which the commodities are bought and sold at
What are the realistic and unrealistic features of the 1962 smith experiment?
+traders ignorant to each other values
+learn about other values by observing their willingness to trade
+small number of trades
-s and d are held constant over trading periods
Results of smith 1962
In early periods actual prices are below the expected price according to supply and demand. In later periods, the actual prices converge to the expected price
How do you calculate the coefficient of convergence for each trading period?
Alpha= 100 x (standard dev of prices around equilibrium)/ predicted equilibrium price
How is the actual efficiency of observed market outcome measure
As a % of the total share actually earned
How does induced value methodology play a role in experimental markets?
Experimental markets rely on the assumption that individuals really are motivated in the way the experimenter intends
How does the slope of supply and demand curves effect the rate of convergence towards equilibrium prices?
Flatter slopes generate faster convergence but steep slopes still converge
How is a double auction market effected when their is only one seller?
Behaviour starts near monopoly price, erodes away, then comes back, then drifts to the competitive equilibrium again. So there is some evidence of monopoly power
Describe the gode and sunder 1993 experiment
-They focus on double auctions
-they compare 3 treatments of human traders, unconstrained computers, and constrained computers
Results of gode and sunder 1993
-human results replicate previous DA convergence results
-unconstrained computers give random results, looks like static
-constrained computers show mild convergence to predicted price, there is no learning through oeriods
What is gode and sunder’s main conclusion from their 1993 experiment?
The primary cause of the high allocative efficiency of double auctions is the market discipline imposed on traders; learning, intelligence or profit motivation isn’t necessary
What is a posted offer institution?
-Sellers set and announce a price before trading
-Price held constant during a given trading period
-Buyers shop in sequence
-Very common institution
What is the aim of the Ketcham, Smith and Williams 1984 experiment?
-Examine properties of PO institutions
-Supply and demand schedules implemented- induced value techniques
-compare PO vs baseline DA
How is a post offer institution implemented by Ketcham, Smith and Williams 1984?
-multiple buyers and sellers
-buyers (sellers) can buy (sell) multiple units
-buyers (sellers) given value (cost) for each unit
-sellers set proposed price p* and quantity willing to sell at p*
-buyers randomly take turns to accept any offers
Results of Ketcham, Smith, and Williams 1984
-DA as expected convergence to price and quantity
-PO shows some convergence but generally price is higher than predicted
-monopolist in PO market gives monopoly prices as expected
-support the conclusion that PO is less competitive than DA
Why is PO less competitive than DA?
When a monopolist sets the price they benefit since it is like an ultimatum game with the surplus area. The seller can take the majority and the buyer will generally still accept the offer
Characteristics of experimental asset markets
-Objects traded has asset like features, value is uncertain, may produce flow of returns over time
-traders
-may be asymmetric info
What is the efficient market hypothesis?
Asset markets are efficient in the sense that market prices reflect underlying values of assets taking account of relevant information
Speculation hypothesis
Asset prices may systematically deviate from underlying values. In particular because the pursuit of speculative profits can lead to prices deviating from the fundamentals
Why is EMH difficult to test in the field?
Because we don’t know the fundamental values or what relevant info is
Advantages of experiments
Experiments can:
-control fundamentals- e.g. implement assets with present value known to experimenter
-implement markets with rational expectations equilibria known to experimenter
-manipulate info structures
Describe the features of the classic asset
market experiment by plott and sunder 1982
-investors are endowed with 2 assets, which pay dividends each quarter, and cash
-they can choose to hold buy or sell
-state of nature is randomly determined in each period
Results of plott and sunder 1982
In early periods allocations more consistent with PI. In later periods, allocations move consistent with RE. Uninformed agents learn not to buy at prices that seem attractive (given expected dividends)
