FocusEconomics had 26 people, including me, answer the question: “How and when will the next financial crisis happen?” Here is my answer:

There are two different types of extreme financial events; one is a crisis, the other isn't. In 2008, banks and other financial firms were so highly leveraged that a modest decline in housing prices across the country led to a wave of bankruptcies and fears of bankruptcy. By contrast, the dot-com crash at the beginning of the millennium led to a large decline in stock prices, but no domino effect beyond that. Because most stock-holding is done with wealth people actually have, rather than with borrowed money, people's portfolios went down in value, they took the hit, and basically there the hit stayed. Leverage or no leverage made all the difference. Stock market crashes don't crash the economy. Waves of bankruptcies in the financial sector—or even fears of them—can. The lesson is: Don't allow much leverage in the financial sector! 

Financial leverage means borrowing a lot. What does it mean to not allow much leverage? It means requiring banks and other financial firms to raise a large share (say 30%) of their funds either from their own earnings or from issuing common stock whose price goes up and down every day with people's changing views of how profitable the bank is. When people buy common stock, they know they are taking on risk. By contrast, when banks borrow, whether in simple or fancy ways, those they borrow from may well think they don't face much risk, and are liable to panic if there comes a time when they are disabused of the notion that the don't face much risk. Common stock gives truth in advertising about the risk those who invest in banks face. One might question whether banks should be forced to issue stock to immediately get up to 30% of their funding coming from stock, but forcing banks to retain all of their earnings until and unless they reach that 30% threshold of being financed by stock has no real downside. 

If banks and other financial firms are required to raise a large share of their funds from stock, the emphasis on stock finance 

1. Provides a strong shock absorber that not only turns defangs the worst of a crisis, and also 

2. Makes each bank enough safer that after a period of market adjustment, investors will treat this low-leverage bank stock (not coupled with massive borrowing) as much less risky, so the shift from debt-finance to equity finance will be more costly to banks and other financial firms only because of fewer subsidies from the government: less of an implicit too-big-to-fail subsidy, less of an implicit too-many-to-fail subsidy, and less of the tax subsidy to borrowing.

My view on this owes a lot to an important book by Anat Admati and Martin Hellwig: The Bankers' New Clothes. This book has persuaded many economists. 

Sometimes people point to aggregate demand effects as a reason not to reduce leverage with "capital" or "equity" requirements as described above. New tools in monetary policy should make this much less of an issue going forward. And in any case, raising capital requirements during times of low unemployment such as now is the right thing to do. Sometimes people think the economy as a whole will take on too little risk if banks are required to have low leverage. My view is that if the taxpayers are going to take on risk, they should do it explicitly through a sovereign wealth fund, where they get the upside as well as the downside. (See the links here.) The US government is one of the few entities financially strong enough to be able to borrow trillions of dollars to invest in risky assets. However controversial that is, providing an implicit guarantee to financial firms that get the upside while the taxpayers foots the bill for the bailouts should be more controversial. The way to avoid bailouts is to have very high capital requirements, so bailouts aren't needed.