Do you want to own growth or value stocks? The obvious answer is “both”.

We all want to own stocks that are a good “value”, priced at a bargain, relative to whatever “growth” the future may have in store. We want companies that offer growth and value. Unfortunately, the investment community uses these words too liberally, with the same words used for far too many concepts.

Wikipedia defines growth as “an increase in some quantity over time”; for an investment, this might mean growth in sales, profits and dividends. Value is “how good a thing is; how much money someone will exchange a thing for”. For an investment, this might mean how much a stock will fetch today, or at some future date.

To the academic community, growth and value have come to have a specific meaning. But it’s not the meaning that most of us might expect. Growth refers to companies trading at higher valuation multiples – often based simply on the price-to-book-value ratio – than the broad market, not companies with superior past or expected growth in sales, profits or dividends. Value refers to companies trading at lower valuation multiples than the broad market, not companies that are cheap relative to past or expected growth.

These definitions are not without merit. They are simple. Also, if a company is trading at a premium to the valuation multiples of the market, of course the market must be expecting faster-than-market growth to justify the higher-than-market multiple.

To a “quant” manager, the academic definition prevails. But, to an old-school investment manager, the basic “Wikipedia” definitions are more relevant. Value is what an investment is judged to be worth, not right now (we know that’s just the price) but at some point in the not too distant future. Growth is what has happened, and may happen, to a company’s business.

Which definition matters? Well, they all do! But it’s a pity that the same words are used for multiple concepts.

It’s troubling that the academic and “quant” communities use the words “growth” and “value” to refer not to growth or value, but to a relative valuation multiple. What’s the company actually worth? How does the price compare with the fair value of the stock? Not relevant: price equals fair value. Are “growth” stocks growing fast or slowly? It doesn’t matter; it only matters that the market collectively thinks they deserve a premium price. Are “value” stocks trading at a discount to their eventual fair value? It doesn’t matter; it only matters that the market consensus wants to see a below-average valuation multiple.

These definitions do not define a type of company, nor do they relate to a company’s past or future “growth” in objective fundamental metrics of company success, except to the extent that a higher valuation multiple tacitly implies an expectation of superior future growth. In fact, academia presupposes that price and value are identical, that markets are “efficient”, and so price identically equals fair value for all assets, in all markets, every minute of every day. In such a world, the expected returns of growth and value stocks are the same, on a risk-adjusted basis.

But, value usually beats growth. Academics attribute this to hidden risk factors, and other academically pure causes, that allow them to hang on to the fiction that markets are “efficient”. Most quant managers then latch on to this “anomaly”, where value usually beats growth, as the simplest and most direct way to add value (another meaning of that same word!) relative to the broad market.

Indeed, this academic definition of growth and value misses something pretty profound. If value usually beats growth, then “value stocks” (defined narrowly as those trading below the average valuation multiples for the market) are, on average, cheap relative to their unknowable true fair value, and “growth stocks” (companies trading above the market valuation multiples) are, on average, overpriced. This would mean that the market usually overestimates its own ability to forecast future growth, and so overpays for high-multiple stocks relative to low-multiple stocks.

It’s easy to imagine that this could be perfectly normal, even if it’s at odds with finance theory: people love to forecast the future by extrapolating the past, so paying premium multiples for companies that have grown quickly is wholly consonant with human nature. It might also mean that there’s a natural spread between growth and value stocks, that the market should pay for perceived future growth differences, and that value will usually beat growth when the spread is wider than this natural range, and growth generally beats value when the spread is too narrow.

What does this mean for all of us, the ordinary investors? We should invest in growth companies only if we think they’ll grow faster than the growth that would be required to justify their high multiples; shun value companies only if we think they’ll grow even slower than the anaemic pace that the current share prices would suggest.

The writer is chairman of Research Affiliates. arnott@rallc.com