I hate to spoil the party, but it's beginning to look like loan growth across U.S banks is too good to be true. Here are some reasons why.

The most evident reason is that banks, once again, are defying the laws of gross domestic product. Think of GDP like gravity. Both are natural forces that determine outcomes and cannot be defied without consequences.

Like banks, the growth of companies like Walmart, Procter & Gamble and Consolidated Edison is largely determined by the overall growth of the economy. But banks can get into trouble when they periodically grow much faster than the economy by booking poorly underwritten loans.

Since 1984 the banking industry's loans have grown annually on average 1.66 times faster than GDP. History shows that when the industry's loan balances grow more than twice as fast as GDP, eventually there is a reversion to the mean. From 2003 to 2007, loans grew 3.4 times faster than GDP. What followed was a violent reversion to the mean that ushered in the financial crisis.

There are signs yet again that banks' level of risk is outpacing broader growth. In 2015, bank loans grew 2.72 times faster than GDP after having grown 2.31 times faster in 2014. In the absence of stronger economic growth, many bankers need to think about curbing risk appetites.

Secondly, a review of 80 different fourth-quarter earnings call transcripts reveals that CEOs were strongly bullish at yearend 2015. Two-thirds of the CEOs were bullish about their bank's current performance and future opportunities. Only three of the 80 executives expressed bearish concerns. As further evidence of the bullish spirit, 18 of the CEOs indicated their bank recorded double-digit year-over-year loan growth. Contrarians might find banker confidence worrisome.

It is worth noting that executives at both BB&T and Fifth Third Bank referenced the GDP growth rate as a reason for caution during their fourth-quarter earnings calls. Good for them.

Third, one of my favorite metrics is the industry's loan-loss provision as a percentage of assets. In 2014 and 2015, this ratio hit historic lows. However, in the fourth quarter the ratio jumped to its highest level since the third quarter of 2013. Although the rate is still much lower than the 32-year average, the trend needs to be carefully monitored as history shows that the loan-loss provision rate is a meaningful leading indicator of future profitability weakness.

The historically low loan-loss provision rate provided a big boost to bank earnings in 2015. But the industry will experience a $32 billion pretax reversal in earnings once the provision rate climbs to just two-thirds of its historic average. Presumably bank regulators are all over this metric. However, even though Federal Deposit Insurance Corp. Chairman Martin Gruenberg – in remarks timed with the agency's fourth-quarter Quarterly Banking Profile – mentioned the recent rise in loan-loss provisions toward a more normal level, he did not reference the fact that provisions are still unsustainably low.

Fourth, just look at the warnings of M&T Bank Chief Executive Robert Wilmers, at 81 years young, writing in the bank's 2015 annual report. I take note when he says that "lending standards have loosened over the last several years." His insightful letter is must reading for bankers, directors, investors, regulators and public policymakers. I am reminded of that banker saying: "There are old bankers and bold bankers but no old bold bankers." Younger members of the industry would be wise to pay attention to what bankers like Wilmers have to say.

Wilmers' commentary about the fragmented nature of bank supervision in the U.S. echoes concerns I documented in "Broke: America's Banking System." The fact is that this nation's regulatory construct is a source of systemic risk as regulators sacrifice the benefit of coordination for the protection of turf.

Fifth, even though my research shows that commercial and industrial lending has provided superior risk-return characteristics over the past decade, in recent years this "trade" has become quite crowded. It appears intense competition for C&I loans is compressing yield, which in turn jeopardizes future risk-adjusted profitability. Getting paid appropriately for risk requires skill and a disciplined commitment to shareholder value.

My biggest lesson learned in banking is that when things appear to be too good to be true, things are indeed too good to be true. My second-biggest lesson is that the industry fails to learn from its mistakes.

Maybe this time will be different. But don't count on it.

Richard J. Parsons is the author of "Broke: America's Banking System." The analysis for this post is from his new book, "Investing in Banks: Strategies and Statistics for Bankers, Directors, and Investors," to be published by the Risk Management Association in April.