We liked it at $10, so we like it even more at $7

7 minutes reading time

It is an axiom among value investors that averaging down is usually the right thing to do. If you though a stock was cheap at $10, then you should be even more keen to buy it at $7. A perfectly logical approach you’re looking for value. Except that the skewed distribution of value stocks means that there’s a significant risk that you’ll buy more of a losing stock. What’s a value investor to do?

Step one is to understand the maths of averaging down. Averaging down only works if subsequent purchases lower the cost basis of the stock below the price at which you sell.  In short, not all methods of averaging down work.

Step two is to understand the risks involved in averaging down. An investor that averaged down is committing more capital to (what has up until that time been) a money-losing stock. We know that the majority of value stocks are losers  and that losing stocks can drive return outcomes for value stock portfolios.

Step three is to try to understand how current market conditions affect the likelihood that averaging down will work. The short answer is they probably make averaging down less likely to workHigher valuations among value stocks reduce the margin of safety.

Topping up doesn’t work!

Topping a losing position back up to its target weight is what most investors mean when say they’ve averaged down. Typically, they’ll start with a x% position and re-balance back to this level as the stock price falls. This is unlikely to work. It fails to lower the cost base of the position far enough to profit from a subsequent recovery in the stock price.

We can illustrate this using Gilead ($GILD) Sciences as an example. In August-2015, the once biotech stock had fallen approximately 20% from its June 2015 high.


$GILD had net cash on the balance sheet and was extraordinarily profitable. The risk of bankruptcy and a permanent loss of capital are virtually zero. The share price was falling because $GILD had effectively cured Hepatitis C. And cured patients don’t need any more drugs. Consequently the market’s concerns over fewer Hepatitis C patient starts drove the stock price lower. This was compounded by other concerns, including political pressure to lower drug prices and competing Hepatitis C and HIV treatments.

Imagine that a portfolio manager decided to invest in $GILD in August 2015. He figures that these issues are short-term. Our imaginary portfolio manager finds himself in the following situation:

  • Portfolio value = $10,000,000
  • Initial position size = $100,000 or 1%
  • Top-up position to $100,000 by buying more shares each time $GILD falls -10, -20, -30 and -40 per cent below the initial purchase price.

The chart below shows what happens. The solid black line shows the principal amount invested in $GILD. The orange line shows the market value of the $GILD position. On the right axis the grey dotted line shows the number of shares held.


Our portfolio manager invests a total of $144,005.07 by topping up the position 4 times:

  1. 107 shares @ $94.16
  2. 190 shares @ $79.86
  3. 85 shares @ $74.76
  4. 191 shares @ $65.40

The market value of the holding is $127,970.00. Our portfolio manager is -$16,035.07 in the red despite a +20% increase in the share price since early June.


Simply topping up the position hasn’t lowered the cost base enough to turn the position into a profitable one. It is still $94.24.

A periodically re-balanced portfolio (e.g. many smart beta strategies) is likely to have similar results. The cost basis will largely depend on the initial purchase price and the re-balancing trades are unlikely to lower the cost base enough for averaging down to work.

The only way to reduce the cost base is to buy the majority of the shares at the low price. Instead of buying a full position at the outset, our portfolio manager should have bought a much smaller position and scaled the position size up as the share price went lower.

Averaging down like you mean it – the Martingale

A martingale is a betting strategy where the size of the bet increases dramatically after every loss. The ideas is that the first win will then recover all previous losses plus a profit. “Double or nothing” is a form of martingale betting strategy.

In this case a falling stock price counts as a loss. The idea is to increase the position size as the stock falls in the hope that when the price rebounds you will have lowered your cost base enough to make a profit.

Imagine a second portfolio manager who decides to follow a martingale strategy. She does the following:

  • Portfolio value = $10,000,000
  • Initial position size = $50,000 or 0.5%
  • Increase the position to $100,000 (1%) when $GILD falls -20 per cent
  • Increase the position to $300,000 (3%) when $GILD falls -40 per cent

The chart below shows her results. the solid black line shows the principal amount invested in $GILD. The dark blue line shows the market value of the $GILD position. On the right axis the grey dotted line shows the number of shares held.

GILD Martingale

As you can see, the portfolio manager in this example starts out small and takes progressively larger positions. She also waits for larger drops in the price before pulling the trigger.

Our portfolio manager invested a total of $329,851.12 in $GILD and made a profit of $54,226.38, more than the value of her original 0.5% portfolio position. To do this, she had to increase the position in $GILD from 0.5% to 3% of her portfolio.

The chart below compares the performance of the “top up” strategy (orange) with the “martingale” strategy. The martingale clearly wins. So why do most portfolio managers top-up instead of double down?


What about risk?

Because the martingale is risky! Here’s why:

  • You now have a multiple of your original principal invested in a stock that, up until this point, has lost you money.
  • This probably means that it’s more likely to be one of the 60-odd per cent of value stocks that lose money.
  • You must have the guts and contrarian willpower to keep doubling down/reducing the cost base of the position.
  • You need to have the capital to keep scaling up the bet.
  • If you’re a fund manager, you must have patient and understanding clients!
  • Averaging down means often owning more of a stock that has difficulty growing its intrinsic value over time.

It also increases risk at the overall portfolio-level:

We can see the potential  in the chart below (again using $GILD as an example) which shows the daily performance impact of topping up a 1% position (blue) versus a martingale strategy (orange) starting at 0.5% and finishing at 3%.

GILD Perf Impact

The martingale strategy starts off with a smaller impact on the portfolio as it’s initially a 0.5% holding instead of a 1% holding. The performance impact is the same when both strategies hold 1% in $GILD. Finally the performance impact for the martingale strategy significantly larger as the position keeps scaling up.

The $GILD example here is a happy ending for the portfolio manager using a martingale strategy. Imagine how she would feel if $GILD continued to lose money?

That’s why fund managers usually stick to topping up in small increments that are unlikely to improve performance. I’ve had many frustrating conversations with fund managers that went something like this:

Me: I see that stock XYZ has fallen 18% today on the back of a earnings downgrade. What are you going to do?

Fund Manager: The analyst is running the revised numbers though the valuation model. We”re also trying to organise a call with company management next week.

Me: OK, so why don’t you sell the stock now, run the numbers, talk to management and if it all stacks up, buy it back next week?

Fund Manager: (Hesitates) We may not be able to buy it back at the same price. Besides it’s likely that we’ll buy more. We don’t want to generate unnecessary turnover.

Me: But you’d be looking at the stock with a fresh pair of eyes and free of any behavioural bias (e.g. endowment effect). Anyway, how much do you think you’ll buy?

Fund Manager: XYZ was a 2.5% position in our portfolio. The benchmark weight is 1.25%, so we were 125 bps over-weight. We might increase it 3%.

Me: Is that all? It’s almost 20% cheaper. Surely you should buy more if you think it’s attractive?

Fund Manger: There’s also risk to consider. We’re not comfortable having more than 3% invested and being 175bps over-weight in a stock this risky.

Me: I’m confused.

The Impact of Current Market Conditions

An earlier post highlighted that cheap value stocks in the S&P 500 are now 30% more expensive (on a Forward P/E basis) than they were during the tech bubble. They are more expensive (on a Price/Sales basis) than they have ever been since 1986.  This is not only true of US stocks. In markets around the world, so-called cheap stocks aren’t that cheap. 

My guess is that this also increases the risk of averaging down. Starting from a level that’s where the average value stock is 30% expensive means that losers now have much further to fall.

The last word

Paul Tudor Jones had a sign up in his office that “Losers Average Losers”.


Somehow, I doubt that most value investors are going to take advice from a trader – even if, on average, it’s good advice.

So I found a quote by Lee Ainslie that contains some helpful suggestions:

One thing I learned from [Tiger Management’s] Julian Robertson is the concept that there are no holds. Every day you’re either willing to buy more at the current price or, if you aren’t, you should redeploy the capital to something that you believe does deserve incremental capital…

… We distribute every day something we call the Sheet of Shame. It shows our ten largest losses, cumulatively from the inception of the position, year-to-date, month-to-date and yesterday. It’s a way of focusing our attention on what’s not working. There are only two ways to get something off the Sheet of Shame – which people are eager to do – either eliminate the position or increase the position and be right, earning some of the losses back.

The key point of this quote is this: inaction isn’t an option. An investor needs to get comfortable that the stock is an even better bargain. If they can’t, there’s no reason to hold the position. Either way, investors need to take action.

Other posts in this series:

  1. Value Investing – Don’t get Skewered!
  2. Losers in the Driver’s Seat
  3. Margin of Safety – the Clock’s Ticking
  4. Reinvestment risk – Got any ideas?
  5. Growth – the Enemy of Value Stocks
  6. When cheap isn’t cheap
  7. Hitting the target
  8. We liked it at $10, so we like it even more at $7
  9. Value Investing – Some Suggestions

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s