Consumer stocks are the new dividend kings according to two new exchange-traded funds that use artificial intelligence to predict which U.S. and Canadian companies will boost their payouts the fastest.

Bristol Gate Capital Partners Inc., a Toronto-based firm with about $1.1 billion under management, developed a platform to forecast dividend growth that’s evolved in the past decade from a “souped-up Excel model” to a machine-learning algorithm supplemented by fundamental analysis, said President Michael Capombassis.

The algorithm looks at roughly 500 factors to determine what drove dividend growth over the past 20 years and to predict what companies will exhibit those factors in the coming 12 months. Bristol Gate’s portfolio managers then apply fundamental analysis to choose the funds’ holdings from the model’s recommendations.

“The core theme behind our strategy, and this is backed up by our research and by our testing, is that if we could identify the highest dividend-growth stocks in the market in the coming 12 months, and we broadly built a portfolio around those stocks, then over time we would outperform the market with lower downside risk,” Capombassis said in an interview at the firm’s offices.

Financials Passe

The resulting Concentrated US Equity ETF and Concentrated Canadian Equity ETF, launched in February, don’t look like most other dividend funds. Bristol Gate’s U.S. ETF shares only two holdings with the S&P 500 Dividend Aristocrats Total Return Index, arguably the best-known U.S. dividend growth benchmark.

“If you buy the Aristocrats product, it’s the assumption that what worked in the past will work in the future, whereas we’re looking for companies that are about to go on a sustained run of dividend growth,” said Jamie Houston, manager of sales development at Bristol Gate.

The Canadian ETF only has about 16 per cent of its holdings in financials, long favored by dividend-seeking investors, versus 35 per cent for the S&P/TSX Composite Index.

“You’re not buying the index with either of these two products,” said Capombassis. He cited the example of CVS Health Corp., which was owned by a Bristol Gate fund that was driven by the same strategy as the U.S. ETF. When Bristol Gate bought CVS in 2009, most of the conventional dividend funds ignored it because its yield was too low. But Bristol Gate’s algorithm identified it as a dividend grower and, sure enough, its dividend rose by 24 per cent annually until it was sold in 2016.

Overweight Industrials

Bristol Gate’s U.S. ETF is overweight industrials and consumer discretionary relative to the benchmark. Its holdings include Home Depot Inc., Starbucks Corp., Honeywell International Inc. and Visa Inc. Since the strategy’s inception as an offering-memorandum fund in 2009, it has returned 322 per cent versus 259 per cent for the S&P 500 Total Return Index, according to Bristol Gate.

The Canadian ETF is overweight consumer staples and discretionary, with holdings like Alimentation Couche-Tard Inc., Canadian Tire Corp. and Maple Leaf Foods Inc. The Canadian strategy, which Bristol Gate been running with its own money since 2013, has returned 81 per cent versus 53 per cent for the S&P/TSX Total Return Index since inception.