OFC. I guess the literature suggests algo collusion less prone to deviation.

Bottom line? The case for rule-based antitrust policy is fairly straightforward. Unfortunately this established wisdom is being forgotten on account of a fixation on business size. Time to reread the consistently classic Kydland and Prescott paper from 1977 😃 10/n

In turn, the risk premium on acquisitions rises, valuations fall, and capital for innovation dries up. Worse, enforcers may extract disproportionate remedies upfront, knowing they can later revisit deals 9/n

If the FTC and US DOJ use discretion to renege on the rule-based policy of non reversals of approved mergers, then the risk is that competitors of acquirers that did not manage to seize good assets use section 2 and the antitrust enforcers as an appeals process 8/n

What is the unintended consequence you may ask? It takes thinking to see one, but it's fairly clear. In digital, there is a high rate of new firm creation and entry. Hence, there’s a vibrant market for corporate control, where firms compete for acquisition of the best assets 7/n

Now, back to antitrust: if the FTC or US DOJ is given discretion to unwind deals years later using antitrust law, merger approval becomes un-credible. Cleared no longer means cleared 6/n

Flood zones are a good illustration: the Gov states a “no bailout" rule, but always pays after a flood. The result? People build in flood zones 5/n

A time inconsistent policy has the allure of efficiency. The false negative is reversed. But econ say that time-inconsistent policies also lead to moral hazard 4/n