Price Discovery in Online Markets: Convergence, Asymmetries and Information
Last registered on March 09, 2018


Trial Information
General Information
Price Discovery in Online Markets: Convergence, Asymmetries and Information
Initial registration date
March 09, 2018
Last updated
March 09, 2018 4:37 PM EST
Primary Investigator
ETH Zurich
Other Primary Investigator(s)
PI Affiliation
ETH Zurich
PI Affiliation
ETH Zurich
Additional Trial Information
Start date
End date
Secondary IDs
Prices tend to converge rapidly to competitive prices in traditional markets for non-durable goods. An open question has been whether and why the same applies in online markets in light of the increase in size, anonymity and information decentrality that characterizes them. We conduct
controlled experiments.

We test the following main hypotheses:

[H1.] Direction of successful/unsuccessful bids and offers are reinforced/reverted.
[H2.] There are asymmetries in terms of adjustment size depending on whether a bid was successful/unsuccessful.
[H3.] The adjustment size depends (positively) on the level of the bid.
[H4.] Information affects the bid/offer behavior. (How does it affect findings regarding H1-H3?)

External Link(s)
Registration Citation
Duran, Diego, Heinrich Nax and Bary Pradelski. 2018. "Price Discovery in Online Markets: Convergence, Asymmetries and Information." AEA RCT Registry. March 09.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
The key dependent variable is the bid/ask that a market participants makes. It will be measures in absolute terms and relative to the agent's willingness to pay/buy, which is a variable randomly allocated by us as the experimenter.

We run two-sided (buyer,seller) trading markets where subjects can buy/sell identical goods (one per trading period). We vary the information content subjects receive about bids, asks, realized prices.

Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
The outputs are:
* explain individual bid adjustments
* explain which information is relevant
* explain asymmetries between buyers and sellers
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Our experiments are conducted on \emph{Scienceexperiment Online} `SciOn' at \texttt{} (please see \texttt{} for a demo). SciOn is our own new real-time trading platform designed to handle multiple large real-time trading experiments in parallel. We developed it from scratch during 2016 using the programming language PHP. Simultaneous real-time play requires instantaneous synchronization, which we ensure via a WebSocket computer communication protocol (instead of the standard HTTP protocol).
This protocol allows interaction between a web client and a web server with lower overheads, thus facilitating real-time message exchanging. Importantly it allows to quasi-constantly `refresh' a users page, thus providing every user with real-time information about other users' bids, offers, and realized prices. In addition, the efficiency of the WebSocket protocol allows running multiple, separate trading rooms at the same time, each of which may be large (we tested our platform successfully for over 10 trading rooms with at least 20 participants each, and for trading rooms of up to 400 participants).
To possess live control of the current status of the whole system (message dispatching and queuing, transaction count, database, query, etc.) we developed a server monitor, which also contains an emergency module that enables communication with the participants in case of failure in the main server.
Experimental Design Details
Randomization Method
Subject recruitment is via Amazon Mechanical Turk. A subject is randomly assigned a position in the trading game by our software.
Randomization Unit
Individuals are randomly assigned trading positions.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
we are planning to recruit ca. 400 subjects into 20 clusters. That is each market will roughly consist of 20 agents, 10 buyers and 10 sellers.
Sample size: planned number of observations
We expect to observe bids and offers in the range of 100,000. We expect to observe more than 1000 realized prizes.
Sample size (or number of clusters) by treatment arms
20 trading experiments witch ca. 20 traders each
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
5% difference (based on
Supporting Documents and Materials

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
IRB Name
DeSciL Review Board ETH Zurich, Board Member Expedite Waiver
IRB Approval Date
IRB Approval Number
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports and Papers
Preliminary Reports
Relevant Papers