I have to disagree with Matt Forstater about what defines "fiat   
currency."  He implies that it is essential that it involve   
liabilities.  But this seems exactly backwards.  Under a gold standard,   
paper money was a liability:  it was exchangeable by right into gold   
coin.  Fiat money differs from earlier paper money in that it is not   
exchangeable by right into something else.  That older, post-gold-standard   
U.S. paper money continued to say "payable to the bearer on demand, the sum   
of . . ." was just an artifact of its prehistory as a parasite on a   
metallic standard.  Recent paper money drops this pretense.  It is   
similarly a pretense for the Federal Reserve to treat its banknotes as   
liabilities.  
  
Yes, the aluminum disks of the capuchin moneys do not involve liabilities,   
but in no practical sense does a U.S. dollar bill either.  But aside from   
that, are the monkey's involved in barter when they use the disks?  Only if   
all exchanges using gold coin were also barter.  One can easily define   
barter in such a way to make that true.  But the moneyness of gold coins   
comes from a social property that is worth distinguishing from   
barter.  Gold can be the source of direct utility (in jewelry or electrical   
circuits), but to most holders of gold coin, it is valuable because it is   
accepted as valuable by other people, and therefore an excellent   
intermediary in transactions.  It is the presence of such an intermediary   
that distinguishes monetary exchange from barter.  The gold standard had   
this property (typical money was not a liability); the gold-(or   
silver-)backed paper money system had it (typical paper money was a   
liability); and the current fiat-money standard has it (typical money is   
not a liability).  The monkeys' aluminum disks presumably are not a source   
of direct utility, but are held because they facilitate exchanges for   
things that are sources of direct utility:  that looks like fiat money.  
  
Kevin Hoover