Notes on "Herd behavior in financial markets: an experiment with financial market professionals" (Cipriani, M.; Guarino, A.)

Herd behavior in financial markets: an experiment with financial market professionals

Cipriani, M.; Guarino, A.

Journal of the European Economic Association , 7 (1) pp. 206-233. (2009)

2009-03

the extent to which trading in financial markets is characterised by herd behaviour 

herding may have both on financial markets’ stability and on the markets’ ability to achieve allocative and informational efficiency

Surveys:

Gale 1996

Hirshleifer and Tech 2003

Charley 2004

Vives 2008

To test herding models directly with data from financial markets:

sample consists of financial market professionals

the existing literature has tested for the presence of herding in a market where, according to the theory, herding should never arise

use a strategy method-like procedure that could help to detect herding behaviour directly

Treatment I: subjects should use their private information and never herd

Treatment II: herding becomes optimal because of event uncertainty

The theoretical model of Avery and Zemsky (1998)

similar to that of Glisten and Milgrom (1985) and Easley and O’Hara (1987)

An informed trader engages in cascade behaviour if he chooses the same action independently of the private signal. If the chosen action conforms to the majority of past trades the trader engages in herd behaviour. If the chosen action goes against the majority of past trades the trader engages in contrarian behaviour.

the challenge for future research is twofold. On the one hand, the existing experimental results offer suggestions for research with field data, which should study whether the behaviours observed in the laboratory are present in actual financial markets. On the other hand, more theoretical work is needed to capture the behaviour that the present model is unable to predict, such as contrarianism and abstention from trading activity.

Notes on "Big Data: New Tricks for Econometrics" (Hal R. Varian)

Big Data: New Tricks for Econometrics

Hal R. Varian

2014, Vol.28(2), pp.3-28 [Peer Reviewed Journal]

the sheer size of the data involved may require more powerful data manipulation tools

we may have more potential predictors than appropriate for estimation, so we need to do some kind of variable selection

large datasets may allow for more flexible relationships than simple linear models

Einav and Levin 2013: new more detailed data

Sullivan 2012, Google uses many of these tools

Out-of-sample predictions:

since simpler models tend to work better for out-of-sample forecasts, machines learning experts have come up with various ways to penalise models for excessive complexity - regularisation

it is conventional to divide the data into separate sets for the purpose of training, testing, and validation.

the standard way to choose a good value for such a tuning parameter is to use k-fold cross-validation

ways to improve classifier performance:

bootstrap involves choosing a sample of size n from a dataset of size n to estimate the sampling distribution of some statistic. A variation is the “m out of n bootstrap” which draws a sample of size m from a dataset of size n>m.

Bagging involves averaging across models estimated with several different bootstrap samples in order to improve the performance of an estimator.

boosting involves repeated estimation where misclassified observations are given increasing weight in each repetition. The final estimate is then a vote or an average across the repeated estimates.

Random forests is a technique that uses multiple trees. A typical procedure uses

the following steps.

1. Choose a bootstrap sample of the observations and start to grow a tree.

2. At each node of the tree, choose a random sample of the predictors to make

the next decision. Do not prune the trees.

3. Repeat this process many times to grow a forest of trees.

4. In order to determine the classification of a new observation, have each tree make a classification and use a majority vote for the final prediction.

spike-and-slab regression, a Bayesian technique

Notes on "Some reflections on financial fragility in banking and finance" Chick, V

Some reflections on financial fragility in banking and finance

Chick, V

8300 defect for UNSW Journal of Economic Issues , 31 (2) pp. 535-541. (1997)

1997

Keynes: investment causes saving, depends for its validity on some fraction of any new, higher level of investment being financed by the banks, because of their ability to finance in excess of saving

Problems: the amount of investment of financed by banks was probably always small, but it is probably shrinking

The system described in The general Theory, with all its threatening emphasis on uncertainty and instability, is a picnic by comparison.

Notes on "The Minimum Economic Dividend for Joining a Currency" (Union Zorzi, Michele Ca’ ; De Santis, Roberto A. ; Zampolli, Fabrizio)

The Minimum Economic Dividend for Joining a Currency Union

Zorzi, Michele Ca’ ; De Santis, Roberto A. ; Zampolli, Fabrizio

8300 defect for UNSW German Economic Review, 2012, Vol.13(2), pp.127-141 [Peer Reviewed Journal]

how the optimality of a currency union depends on whether it brings an economic dividend in terms of potential growth and the Balassa–Samuelson (BS) effect

Kenen 1969; McKinnon, 1963; Mundell, 1961:

the theory states that the benefits from lower transaction costs and greater price transparency must outweigh the cost of giving up an independent monetary policy and flexible nominal exchange rates

The traditional theory, however, focused mainly on the factors affecting the cost of renouncing to an autonomous stabilisation policy. It did not give prominence to the benefits of a common currency over and above transaction cost savings and greater price transparency, nor on the fact that many of the conditions cited in favour or against the creation of a common currency are not static but endogenous to the policy regime.

Our approach offers the advantage that is conceptually simple and being able to reconcile old and new arguments in favour and against monetary union in a unified framework. 

The main contribution of this paper is to show in a unified analytical two-sector, two-country general equilibrium model, how the optimality conditions for forming a currency union depend on both the increase in potential output and BS effect.

The pros include the removal of the distortionary effects on potential output arising from currency risk and output fluctuation. The cons include the emergence of cross-country inflation differentials due to large intersectoral productivity gaps and the reduced macroeconomic stabilisation in an environment of supply and real exchange rate shocks.

The results suggest that both the BS effect and the size of real exchange rate shocks matter to evaluate the optimality of accessing a currency union.