Here’s my prediction of what happens if Greece votes No

Well, it's not really a prediction, just my best guess about both sides’ next moves and the considerations they’ll be taking into account. Like most guesses about the future, it’s probably wrong, but hopefully illuminating. (If there's a Yes vote, I have no idea what will happen, except that Varoufakis will resign and the Eurogroup offer will be signed.)

***

Immediately after the No vote, Greece demands that the ECB restore full liquidity to the banking system (as any normal lender of last resort is supposed to do). A threat is made -- either publicly stated or implicit but communicated to the Eurozone authorities -- that if this doesn't happen, Greece will immediately issue a parallel currency redeemable against future tax payments.

At that point the ECB has to decide what to do. It won't make the decision without clear guidance from the political authorities, because the issuance of a parallel currency is a major step -- albeit potentially reversible -- towards a Grexit.

So the EU will have to decide which outcome is least unpalatable to it. Of course, neither is desirable from its point of view. If it complies and restores ELA, the bank panic ends, cash controls can be lifted, and a calm atmosphere can proceed in which Syriza can negotiate for a better deal -- now armed with a democratic mandate and a public admission from the IMF that the existing deal on the table was not sustainable.

Obviously that would be a terrible outcome from the EU's perspective. It would be perceived (rightly) as a major political victory for Syriza.

So the EU might refuse to restore bank liquidity. In that case Greece will issue the parallel currency.

In my view, the best way to do this is in the form of tradable tax credits redeemable starting in, say, a year. (See here and here.) A fresh batch of these would be allocated immediately to citizens and firms. These credits are obviously worth something: every retailer can use them to pay his VAT, every individual can use them to pay his payroll tax, etc. (Greek businesses have to pay VAT tax every three months, so these credits will come in handy.) Since they're tradable and valuable, Greeks will be willing to buy these credits for euros, albeit at a discount, mainly reflecting the risk that the drachma will be introduced at some point and the tax credits redenominated. As a result, the credits would be a form of money whose supply would be under the Finance Ministry's control. The result, if it works the way it's supposed to, would be Greece's ability to stimulate aggregate demand and increase economic output, which it can't do as long as the ECB has a monopoly over issuance of means of payment. In Milton Friedman’s terminology, the tax credits would accelerate the velocity of euros inside the Greek banking system.

There has been some talk about the technical and logistical difficulties of quickly changing over Greece’s electronic payments system or distributing currency to ATMs. But as I see it, no such complicated operations are needed. Greece can mail every household a paper check worth, say, 600 euros of future tax relief. Individuals can take the check to a currency exchange [SEE UPDATE BELOW], like the ones at the airport, and exchange it for, say, a 300 euro check, which they can then deposit at their bank. (Banks are closed for withdrawals but they’re happy to take deposits!) 

At that point, 300 real euros will be transferred in the usual way, electronically, from the currency exchange’s (Greek) bank to the customer’s (Greek) bank, and 300 euros will be credited to the customer’s account. The individual can spend the money using a debit card -- debit cards are working normally for domestic transactions -- or make (limited) currency withdrawals. Afterward, the currency exchange can sell the tax credits to business and individuals. Again, the point is that the velocity of money is increased, which increases GDP. And Greece can print as many of these credits as it thinks prudent.

So the EU's decision about whether to comply with Syriza's post-referendum threat will depend on how it views this parallel currency scenario: is it better or worse, from its point of view, than the Syriza-negotiating-triumph victory?

Of course, the upside of the parallel currency for the EU is that it doesn’t hand Syriza a major immediate victory. The obvious downside is that it would clearly be a big step towards Grexit. Moreover, it's a step that allows Syriza to keep its promise to voters not to take Greece out of the eurozone: there would still be euros in Greek bank accounts and the Bank of Greece would still be hooked up to the Eurosystem payments network.

The EU has put on a brave face about not really caring about Grexit, but behind the scenes it is deeply divided. Many on the Right, in Germany and Northern Europe generally, seem OK with the idea. (In fact, Schaeuble himself recently mentioned the possibility of a parallel currency in Greece.) But many others, on the center-left and in Southern Europe, privately view the prospect with horror. Francois Hollande, in particular, is now panicking. All along he assumed that Germany would never push things this far; he thought that if he privately and politely urged Berlin to go easy it would listen to him. Now the masks have come off and France is scrambling. God only knows what Renzi et al are feeling.  

So if  the EU takes this path -- if it denies Greece bank liquidity and forces it to introduce a parallel currency -- the immediate outcome would be a political crisis within the Franco-German core the likes of which haven't been seen in many decades.

Even worse is what might happen after the immediate crisis. If a major expansion of the effective Greek money supply does what one would expect it to -- stimulates the Greek economy -- this would be a real nightmare for the Eurozone, for reasons that are too obvious to explain. In many ways, it would really be the worst of all possible worst-case scenarios, politically speaking. And economically speaking, there is the question of what the markets' reaction would be in Spain, Italy, et al., which until now have weathered the Greek crisis OK.

Of course, the eurozone could retaliate against Greece and shut off its access to the payments network, or achieve the same thing by drastically reducing ELA, thus kicking it out of the euro. Politically speaking, this would presumably require a unanimous vote of the EU heads of state at the European Council. (If the ECB took this step over clear French opposition, I think the European project would be effectively over, at least for many years.)

Obviously it would be terrible for Syriza (and the whole country) if Greece were forced out of the euro. It would cause an appalling economic collapse, a visible humanitarian crisis in a NATO country. But in a sense it would also let Syriza off the hook: Hey, we tried our best to fight austerity within the euro, the voters agreed with us, and then the evil Europeans kicked us out.

So a lot depends on three things, in ascending order of importance:

(1) how smoothly Greece can roll out the currency issuance;

(2) how much it would stimulate the economy;

(3) above all, the Europeans' perceptions of (1) and (2).

The risks are high for both sides, but I think Greece is in a stronger position than most people think. Again, I’m probably wrong.

UPDATE:  Actually, this could be done without middlemen. Banks could accept and deal in the credits directly.