Recently I examined the top-performing bank stocks going back to 1980. That exercise uncovered a few critical characteristics that separate the best banks from the rest.

So let's go over those key qualities of top-notch banks and use them to construct my perfect bank for the next two decades.

The business model
A consistent and simple business model drives the best-performing banks. It does not matter if the bank has $1 billion in total assets or $1 trillion; simplicity reigns supreme.

These banks accept deposits, they make loans, and they provide simple but essential value-added services for their clients. They don't rely on proprietary trading or complex derivative-market speculation.

The best banks collect income from lending and value-added services, and they ruthlessly manage expenses.

Competitive lending
Income from lending is called net interest income. Simply put, it's the difference between what a bank pays on deposits (its cost of funds) and what it collects from the interest on loans. This difference is known as the "spread."

Because banks are all competing to lend to the best customers, they will lower the interest they charge in order to draw those borrowers. That's just the nature of capitalism.

The challenge for banks is to price their loans in such a way that they win the deal and turn a profit. That's why the perfect bank for the next 20 years will have a low cost of funds. Wells Fargo's (NYSE:WFC) low cost of funds is often cited as a key reason why Warren Buffett has invested so heavily in the stock over time. The lower cost of funds allows the bank to aggressively price loans without sacrificing its spread. That's a major competitive advantage, and it's a must-have in my perfect bank.

My perfect bank would make roughly 51%-65% of its total income through these lending operations. This would ensure the bank's focus is on banking, not on chasing the current fad (e.g., subprime lending) or trying to goose short-term returns by sacrificing prudent risk-management.

Diverse income streams
My perfect bank will have profit flowing in from a number of sources other than interest. In addition to fee-driven income -- from fees on deposit accounts, loan origination fees, and so on -- my bank would have auxiliary businesses separate from lending that add income without materially increasing the bank's overall risk profile.

That could be a wealth management business, like Wells Fargo Advisors or Bank of America's (NYSE:BAC) Merrill Lynch unit. Or it could be an insurance unit or trust management business, like those offered by regional player BB&T (NYSE:BBT)

These kinds of products and services naturally complement the core banking needs of commercial and individual clients. A business with a working capital line of credit and a few deposit accounts likely needs a cash management solution. A high-net-worth individual needs wealth management, in addition to a mortgage, checking, or savings account.

These secondary income streams help the bank to pursue those lower-priced, higher-quality loans mentioned above. In other words, management can hit profitability targets without having to chase higher-yielding but riskier loans.

For me, businesses such as market making, trading, hedge funds, and private equity are out. Some companies can succeed in those businesses, but I just don't want my perfect bank stock going down that road. I value simplicity and prudence in a bank, and over the long term, the market does as well.

These noninterest income streams would represent the balance of the bank's revenue in my perfect scenario -- somewhere in the range of 35%-49% of total revenue.

Expenses matter, but maybe not in the way you'd expect
To get from revenue to profit, we have to account for the bank's expenses. Those include employee costs, branch costs, and others. The most important costs to me, though, are loan losses.

I like the efficiency ratio as a tool to measure a bank's cost structure and as a proxy for the institution's credit culture. The lower the efficiency ratio, the more efficient the bank; a bank with a 50% efficiency ratio requires $0.50 to generate $1 of net revenue.

My perfect bank stock will have a respectable efficiency ratio, but it needn't be industry-leading. After all, Wells Fargo has a mediocre ratio, yet it was among the best-performing bank stocks since 1980

It's also important that the bank demonstrate a consistent efficiency ratio over time and through multiple credit cycles. Consistency shows that the bank makes the kind of loans that are repaid, in good times and bad. And while some expenses, such as employee costs and branch expenses, are pretty much fixed, the bank's loan losses are quite variable.

As the following graph shows, return on assets and loan loss provisions have an inverse relationship, which shows the importance of quality loans.

More than anything else, this trait separates the best banks from the rest. That's why this is the most important characteristic of my perfect bank.