Should you average down or up?

One of the most contentious topics in investing seems to be the issue of whether you should average up or average down. That is whether you should consistently add to a winning position or add to a losing position. Great investors hold strong and often contradictory stances on this topic and their views are widely quoted. E.g.

We like to buy stocks which we feel are undervalued and then we have to have the guts to buy more when they go down. Walter Schloss

‘Always sell what shows you a loss and keep what shows you a profit’. Livermore

...a price drop [is] as an opportunity to load up on bargains from amongst your worst performers…a price drop in a good stock is only a tragedy if you sell at that price and never buy more. Peter Lynch

Don’t garden by digging up the flowers and watering the weeds. Warren Buffet

One of the mistakes investors can make when reading these quotes is to read these as strict rules. They often miss the caveats contained within them. For Schloss averaging down has to be in something still ‘undervalued.’ Lynch says a price drop in ‘a good stock’ is an opportunity not a price drop per se. Knowing Buffett’s strong focus on the performance of a business not the market price I’m sure his ‘flowers’ are well performing businesses with good economics and his ‘weeds’ are badly performing businesses. I doubt any of these investors are recommending taking action purely on price action alone yet often we interpret them as such.

The other issue with saying one is ‘averaging down’ or ‘averaging up’ is it suggests that your current average buy price matters. It does not. To the fully rational investor a historic trade price is completely irrelevant. A rational investors asks themselves ‘given all current information do I have the right position size in relation to the risks and potential return?’ not ‘should I average up or average down?’ (See my blog post on portfolio optimisation)

Rules like ‘never average down on a losing position’ can be used to try to overcome behavioural biases like loss aversion but they should be recognised for what they are – an attempt to overcome one’s personal bias – not a general investing rule that everyone should follow. Rules can be very useful way of addressing behavioural biases (see: my blog post on portfolio rules) But for a rule to be relevant then it must address a bias that has led to past under-performance. If you have a tendency not to recognise when the fundamental investment case has significantly deteriorated and have a history of adding to positions that never bounce back then set a ‘don’t average down’ rule. If you have a habit of being overconfident and over-sizing your winning positions then set a ‘no averaging up’ rule.

For everyone else simply optimise your portfolio regularly based on the latest information available and forget what the average buy prices of your portfolio constituents are.