Project Planning

From edegan.com
Revision as of 19:21, 19 July 2012 by imported>Ed
Jump to navigation Jump to search

To Do List

Each project has its own 'to do', but in order to explore new projects or break existing (inc. planned) projects into better questions there are the following tasks to be done:

  • Determine whether there are existing models for the effects of the winner's curse under asymmetric information
  • Determine the extent of the literature on the 'choice' between IPO and acquisition for startup firms seeking commercialization financing

Acquisitions Paper

Status: Old paper needs repurposing, perhaps into multiple papers.
Co-authors: Jim Brander
To do: Lit reviews
Project Quality: All reasonably high
Project Page: VC Acquisitions Paper 

The original paper (Brander Egan (2007) - The Role of VCs in Acquisitions) had a full data rebuild, detailed on the VC Acquisitions Paper page.

We now have a number of possible papers that we could write:

  1. The effects of VC on the returns to acquisitions
  2. The role of Information Asymmetries (IA) in acquisitions of privately-held firms
  3. Measuring IA

(1) could be a JFQA Comment, describing how we get a different result from Masulis and Nahata (2011). This requires that we reconsile Masulis and Nahata's results (and perhaps Gompers and Xuan's too).

(2) is probably the simplest paper to write. We will use continuous measures of IA, and perhaps compute new distance measures.

(3) is best answered using a dataset on acquisition of public firms by public firms (i.e. really Mergers). Also need Dierkins' other measures...

All three papers need lit reviews doing first.

Patent Pools Licensing Rate Paper

Q. Are licensing rates lower for patents included in pools?

Status: Infancy
Co-authors: David Teece, Ed Sherry
To do: Conference call on how to proceed
Project Quality: Depends on data and empirical execution
Project Page: Patent Pool Licensing Project

Mexican-Standoff Paper

Q. What are the equilibria in patent litigation between competitors?

Status: Conception
Co-authors: David Teece
To do: Lit Review
Project Quality: Potential Job Market Paper
Project Page: None yet

Background

Nokia sues Apple, Apple sues Sumsung, Samsung sues Nokia, and so on. There's been an endless round of Patent Litigation between the mobile phone and tablet device manufactures of late. Could this be related to the nature of competition in the markets? Pharma firms apparently sue at every infringement, whereas large semiconductor incumbents apparently never sue each off (are they in a Mexican standoff?). Infringement is obvious for pharma firms and potentially non-obvious for semiconductor firms. NPEs always sue, but they have no production rents to worry about. Some device manufacturers try to shut each other out of the market with ITC injunctions, and sue weak firms...

Description

The project would try to describe at least some of the behaviour above using game theory (probably an infinitely repeated, full information game) and then use litigation data to show that the game models reality well.

The Value of Patent Citations for Startups

Q. Do some patent citations increase value and others decrease it? And can we use renewals to crudely cut between the two?

Status: Advanced
Co-authors: poss. Bronwyn Hall
To do: Rebuild data incorporating renewals
Project Quality: Potential Job Market Paper?
Project Page: None yet

This project might also need data on priority dates and other extentions, though this might be able to be deduced from renewals data. The project either works or doesn't once the data is done.

Technology substitution and the choice of commercialization financing

Q. Do firms with more radical innovations seek initial public offerings?

Status: Emprics is there. Needs a model and writing.
Co-authors: poss. Jim? 
To do: Work on the model, decide whether the paper is feasible.
Project Quality: Potential Job Market Paper? Has a big maintained assumption unless done with two models (and then has a measurement problem).
Project Page: None yet

The maintained assumption would be that citations represent substitution. Otherwise, would need two models (one for subs, one for comps).

There are two possibilities here:

  1. Build out my 'learning' model. This might also give rise to a spin-off model, particularly if I can get complements in there.
  2. Use an off the shelf model (e.g. ?)