Discussion of Gresham's Law by Nobel Winnter Robert Mundell of Columbia University
If you think Gresham's Law is "Good money drives out bad," you'd be wrong. This isn't an easy essay, but it's worth the effort. I'm pasting the section about hoarding just to get everyone in the mood.
http://www.columbia.edu/~ram15/grash.html
Uses and Abuses of Gresham's Law in the History of Money
11. The Evidence from Hoards
The study of hoards is one of the most interesting if recent applications of Gresham's Law. Ever since coinage was created, coins have been collected and sometimes buried as hoards, which resurface for the enlightenment and controversy of later generations of archaeologists, numismatists, historians and economists. Sometimes the finds have been merely spectacular(62) whereas otherwise--like the great discovery of electrum coins in the Temple of Artemis at Ephesus--they have yielded priceless historical information.
Hoards found in the ground have typically been those that have been buried in the past with the intention, but not the realization, of future recovery. Three questions deserve to be asked: (1) Why are hoards created in the first place? (2) Why are they buried? (3) What determines their composition? We can answer these questions in turn.
(1). Collections of coins may exist for a number of reasons, including numismatic interests; Augustus Caesar, for example, was an avid collector. Nevertheless, the main function of hoards is as a store of value, a form of saving, which reflects a desire to preserve wealth for future use. Moreover, hoards are typically a form of liquid wealth, ready cash that is available for use in contingencies. Hoarding always reflects a desire to sacrifice current consumption for the prospect of future consumption.
(2) Hoards are buried because of a real or imagined threat to security. The best proof is that hoarding intensity--measured by the frequency of hoards found--increases during civil wars. In Roman history, for example, hoards increased dramatically during the Second Punic War, c. 218-206 BC, and then again during the Social War, the Civil War and the Spartacus revolt, c. 90-71 BC, and once again during the Civil Wars, c. 50-31 BC. A similar pattern existed in Britain where hoarding frequency soared during its Civil War period c. AD 1625-49. It is probable that this pattern will be confirmed in similar cases.
(3) The composition of hoards is determined partly by Gresham's Law. Assume that the money supply of a country consists of both overvalued and undervalued coins. Let us now suppose that hoarding intensity increases. By Gresham's Law it is the undervalued coins that will disappear and the overvalued coins will remain.(64) The profit motive will ensure that the best coins end up in hoards.
An effective application of Gresham's Law to the study of hoards was made by Sture Bolin in a pioneering work on the Roman monetary system.(65) Drawing on work spanning more than two decades, Bolin published in 1958 a comprehensive study of 150 coin hoards from the period AD 69 to approximately AD 250, found in widely separated parts of the Roman Empire and containing over 82,000 denarii from the Republic and Empire up to the year 193. It is of course well known that the denarii was progressively debased over this period, and this fact should have affected the proportion of old to new (depreciated) coins in the collections. Among his striking findings was the vast number of coins from the Republic in the hoards collected in Germany, the Balkans and Britain, confirming what Gresham's Law would predict, that the outer parts of the empire preferred the relatively undervalued denarii.(66)
The composition of hoards sometimes offers clues about circulation. Other things equal, a large preponderance of coins is, say, the reign of Hadrian or Vespasian would suggest a large production. On the other hand, a low frequency of a ruler's coins does not necessarily imply low production; if it were overvalued, it would not be a suitable candidate for hoards. The worst coins circulate, the best coins end up in hoards.
Given the frequency of hoards discovered, which number in the thousands, empirical analysis can be brought to bear, yielding rigorous statistical inferences that can be used to test the presence of absence of Gresham's Law.