License article

Beware investors: Falling knives really do cut

Show comments

Faced with two competing narratives - "Stocks are on sale!" and "Never try to catch a falling knife" - investors probably ought to keep their hands in their pockets.

Global stocks have fallen sharply, with many stocks in sectors such as finance and energy down by 30 to 50 per cent or more from recent peaks.

That kind of "discount" can act on investors like a pre-holiday sale, getting them interested in securities they might not otherwise consider.

Certainly, rapidly falling markets and plunging company shares are subject to sharp reversals. On Wednesday Deutsche Bank, which had fallen more than 50 per cent from its 2015 peaks, rose 10 per cent from its Tuesday low after reports the company plans to buy back bonds. Similarly, mining shares in the FTSE 350 index, which fell as much as 60 per cent since 2015 highs, recently jumped 25 per cent in less than three weeks.

Buying shares that have been hit hardest is not a winning strategy, according to a review of 25 years of data from Societe Generale.

"Rarely do those stocks that lead us down into a slump provide the best performances on the return back up, and as such investing in 'falling knives' is a bad idea," Andrew Lapthorne, quantitative analyst at SocGen, writes in a note to clients.


Looking at all FTSE All-World stocks since 1990, Lapthorne created baskets of those that have suffered falls from their 12-month peak of 20 per cent, 30 per cent, 40 per cent and 50 per cent. The returns going forward are not encouraging. Note too that many arguments that advocate buying the big early losers are based on bad data that comes with using the current members of a given index, as the truly worst of the wounded ducks will either get kicked out of the index or, gasp, go under.

Taking a closer look, SocGen calculates that the four baskets of plungers, if rebalanced monthly, while having some periods of outperformance, all go on eventually to underperform the market. Looking at how these baskets of fallen stocks do if held for three, six, nine or 12 months, there is consistent underperformance versus the market. This lagging performance is seen across all baskets, from the 20 per cent fall group to the 50 per cent fall group, and across all holding periods.

The general trend is that the underperformance gets worse the longer the stocks are held. For example, after 12 months the basket of shares that originally fell 50 per cent has underperformed the market by six percentage points.

Worst of the worst 

This isn't just a matter of buying into horseshoe stocks just as the car gets popular. While the portfolio of fallen stocks will inevitably be biased toward certain sectors, such as mining and energy now, the stocks that fall the most perform worse over holding periods of up to a year against other shares in their sector. They also underperform their country peers, as well as the broad universe as a whole.

The worst of the worst don't even do well, in relative terms, once market sentiment turns.

"As expected, when markets move downwards, our 'falling knives' portfolio sees even worse performance with an excess return of (about) -17 per cent and a hit ratio (i.e. percentage of periods that the portfolio has outperformed the market) of close to 0 per cent on a forward 1-year basis," Lapthorne writes.

"Perhaps more surprising is the performance of our portfolio in up-markets, where we expect our portfolio to perform better and produce strong relative performance. Instead, while overall the excess performance is positive, it is only by less than 2 per cent, with hit ratios of around -50 per cent."

Which is not to say that all stocks that have fallen by 50 per cent should in all circumstances be untouchable. Selecting based on valuation can help and add genuine value, as can buying in a "cheap" equity market.

It is also not to say that big fallers never recover. Amazon and others are testament to the fact that they do. But looked at as a group, even those that fell sharply during the bloodbath commencing in 2008 tended to fall further, sometimes by another 60 per cent, and largely lag the market on the way back up.

One other factor to consider here: the unpredictability of policy. The results in the study will have been influenced by attempts by central banks and government to prop up banks, both in the Great Recession and during the Asian and even Mexican debt crises.

That policy may yet come back into play, as indicated by current rumours that the European Central Bank may consider buying bank shares. On the other hand, it may not.

Just as during Black Friday sales, investors are probably better off not looking for bargains right now but relaxing with their families.