As the U.K.’s vote to leave the European Union shows, the act of one nation could have widespread implications around the world. Similarly, the U.S. housing market meltdown of 2007 to 2009 led to a global financial crisis the effects of which are still being felt today. These examples underscore the need by nations to pay attention not only to what goes on within their borders, but also to factors that lead to systemic risk that could bring down outside economies.

Traditionally, the response of banking regulators has been to issue microprudential rules, which seek to ensure the safety and soundness of individual banks such as mandating minimum capital levels. If banks are prevented from taking risks, the theory goes, then one avoids risks to the financial system. “The problem with that is it ignores systemic risk,” said Franklin Allen, a professor of finance Wharton and also at the Imperial College London, at a recent conference on risks, regulations and financial stability in Philadelphia. The event was sponsored by the Federal Reserve Bank of Philadelphia, the Wharton Financial Institutions Center, Imperial College Business School and the Journal of Financial Services Research.

While the full facets of systemic risk are still unclear, Allen pointed to six things that lead to it: banking panics; banking crises due to falling asset prices; contagion; financial architecture; foreign exchange mismatches in the banking system and behavioral effects from ‘Knightian uncertainty’ — or uncharted waters such as the recent referendum on the ‘Brexit’. His paper, “The Interplay Between Financial Regulation, Resilience and Growth” discusses some of these causes and offers potential solutions.

Banking Panics

Banking panics in the past two centuries occur when depositors withdraw more money than they need for consumption because they see other people taking money out in a bank run. In the past half century, the solution has been to offer deposit insurance or other guarantees. But the game has changed, Allen said. Today there are many more kinds of deposits that fall outside these guarantees, as well as wholesale funding (money tapped by banks outside of traditional customer deposits) and shadow banking (loosely regulated financial firms such as hedge funds).

“The problem with [focusing on microprudential rules] is it ignores systemic risk.”–Franklin Allen

Moreover, the idea that government guarantees should expand and cover these new instruments to prevent crises doesn’t work. Allen pointed to Ireland as a case example. When its real estate bubble burst, the banks ran into peril. But because the Irish government had a blanket guarantee of bank debt, it also neared collapse and had to be bailed out. “So we need, in my view, a much richer view of systemic risk,” he said.

Asset Price Declines

Asset price declines such as in housing or stocks can lead to bank crises, which could trigger systemic risk. According to Allen, reasons for falling asset prices include the bursting of real estate and other bubbles, a rise in interest rates, sovereign default, mispricings due to limits to arbitrage and ‘flash crashes,’ the business cycle and politics. He said real estate is the culprit in a majority of financial crises — as many as two-thirds of the time. And when the bubble bursts, its effects spread beyond the financial sector.

Loose monetary policy and excessive availability of credit — meaning it is easy and relatively cheap to get loans — create these real estate bubbles. “By lowering interest rates significantly below the current rate of house price appreciation, the Fed effectively created a profitable opportunity to buy property,” he wrote. “Loose monetary policy is arguably one of the main causes for the emergence of bubbles.” Does that mean the solution is to raise interest rates to pierce bubbles? Allen argued that such a blanket solution in a diverse society could backfire. For instance, higher rates could cool housing prices in California but might trigger a recession in Michigan, where home values have not risen as much. Even in homogenous nations such as Sweden, there is a big debate about whether or not hiking rates is the way to go.

Instead, governments could institute more accountability for central banks or issue a clear mandate that they should prevent asset bubbles. For instance, the Fed has a dual congressional mandate of aiming for maximum employment and stable prices by controlling inflation. “Central bankers … need to focus more on financial stability issues, which require a whole range of policies, and a little bit less on whether we have 1% inflation, or 2% or 3%,” Allen said.

“Loose monetary policy is arguably one of the main causes for the emergence of bubbles.”–Franklin Allen

While interest rates globally are at all-time lows, Allen sees them going back up. If it happens gradually, people can adjust to it. But if inflation takes off and rate hikes come quickly, it will cause problems for many assets, even best-rated ones such as German bonds or U.S. Treasurys. “They’re going to fall in value” and cause “potentially huge financial stability problems.”

Besides interest rates, other levers to cool down a housing market include lowering the limits on loan-to-value ratios as property prices go up at a faster rate. This ratio reflects the size of a real estate loan vis-à-vis the value of the property; the higher it is, the riskier is the loan. But Allen said adjusting the ratio may be difficult for commercial property because businesses can use pyramids of companies to increase their indebtedness. Other options are to increase property transfer taxes as housing prices heat up and to restrict real estate loans in certain regions.

Contagion and Uncharted Waters

Contagion occurs when the distress of a financial institution infects others in the system and leads to a systemic crisis, such as what happened in the Great Recession of 2007 to 2009. Allen said central banks often use the risk of contagion as justification to intervene, especially when the financial institution is big or occupies a key position in particular markets. “This is the origin of the term ‘too big to fail’,” he wrote.

While the U.S. has rebounded quite well from the financial crisis as measured by GDP growth, some other economies had not. Allen cited Japan and Finland as examples. Even though these countries’ banks were not exposed to the U.S. housing crisis, their GDP had fallen by 10%. “We don’t understand that,” Allen said. Such uncertainties include the impact of Brexit. “If it’s a vote to leave, direct effects of that would be significant, particularly in Europe,” he said ahead of the June 23 referendum. Indirect effects of Brexit would be a signal to other nations to leave the EU.

While regulations are one way to deal with systemic risk, Allen concluded, “regulation alone is not sufficient to create financial stability.”

How the Banking System Got So Complicated

The rise of international banking crises led to a global, voluntary agreement in 1998 among banks to institute common standards to strengthen their businesses, called the Basel Accord. It was a response to a shared view that the world’s banking system has become riskier as banks’ ability to absorb losses has declined, Wharton finance professor Richard J. Herring said during the conference. The rationale for a global agreement came about because individual countries found out that when they tried to raise capital requirements on their banks, these institutions would shift activities to other nations with laxer rules.

Basel III was a “virtual blizzard of new legislation and rulemaking.”–Richard J. Herring

The Basel Accord — or Basel I — defined the regulatory capital or reserves that banks needed to hold. The goal was to make banks hold higher-quality and more liquid assets so they can better withstand economic shocks. But a lack of consensus between the German and French led to a compromise that resulted in two types of regulatory capital, Tier 1 and Tier 2, Herring said. Tier 1 comprised mainly equity, which conforms to the German view, while Tier 2 included other types of assets such as hybrid instruments and subordinated debt, closer to the French view.

But banks intent on increasing their exposure to risk – which promises higher rewards – had several options to do so without needing to raise required reserves, thereby undermining the intent of Basel I. This led to Basel II in 2004, which added “considerably more” complexity, Herring said. For example, the computation of regulatory capital requirements now entailed more than 200 million calculations, defying effective monitoring by regulators and the market and hampering comparisons across banks or even of the same bank over time. Moreover, new types of securities – or innovations in hybrid capital – degraded the quality of Tier 1 capital.

One product that arose from this complex system was the collaterized debt obligation, or CDO, which became the bane of the financial crisis. A CDO contained layers of debt, such as subprime mortgages, and each could have about 750,000 mortgages with 30,000 pages of accompanying documentation, Herring said. The CDOs were so complicated that their market value became divorced from any rational value based on their fundamentals. When investors began questioning the value of CDOs, markets seized up as a result.

In the aftermath of the financial crisis, Basel III came about, which introduced even more complications. Herring called it a “virtual blizzard of new legislation and rulemaking.” While the third accord tightened the definition of regulatory capital, it multiplied the number of ratios to calculate. Herring argued that a “simple leverage ratio” outperformed more complex risk-based ratio in determining bank strength. Mindful of the criticisms, the Basel Committee on Banking Supervision created the Task Force on Simplicity and Comparability in 2012 to avoid complexity. But Herring said there is “scant evidence” that the group has made an impact.

The problem, Herring said, is that most reform proposals focus on tinkering around the margins of the existing structure — and perhaps the focus should be on fixing the process itself.