Are Your Stop Losses Keeping You Safe?

How sure are you that your stop losses are reducing your risk?

Like what you're seeing?
Try our FREE 30 day trial And test your strategies against historical data immediately

Written by Liam Flavelle on 22 August 2017

  • We analyze the effects of stop losses on the returns of two successful strategies.
  • Applying stop losses negatively affected our models in most cases, with only a handful of settings providing any benefits.
  • How do you know if your stop losses actually work?
What with so many facets to look at and brood over when weighing a stock buy, it's easy to forget about the little things. The stop-loss order is one of those little things, but it can also make the world of difference.
(Source: Investopedia)

We have all probably used stop losses at some stage in our investing careers. We use them to reduce risk, to give ourselves a safety net when we invest and to help us sleep well at night. Even more importantly, a stop loss order takes the emotional influences out of the equation - it triggers at the level you set it to, and doesn't hold on for another few days just because it has a hunch that the stock will bounce back soon.

However, how do you know that your stop is actually reducing your risk? I have analyzed the strategies that were designed in my two most recent Seeking Alpha articles ("Building A Better Value Strategy" and "Building A Low-Risk Dividend Strategy") to see if they could be improved using stop losses.

As usual, I have used the platform to backtest the models - you can find lots more statistics and returns information, together with actual position data, if you click here. This shows you the first strategy tested (the Dividend baseline). To view the other versions detailed here, just click on the History option on the left menu and select the desired strategy version to view.

The Models

As a quick review, our Value strategy gave us the following returns over a 17-year period:

Value Baseline

The key areas to look at in the model are the annual return (CAGR), the max drawdown (Max DD, or the largest peak to trough fall in the value of your holdings) and the risk-adjusted returns (represented by the Sharpe and Sortino ratios). The model's volatility, measured by the fluctuations in its maximum drawdowns, can be seen below:

Value Drawdowns

Our Dividend model baseline is as follows:

Dividend BaselineDividend drawdowns

You can see both models had fairly similar drawdown profiles, with the maximum being achieved in 2008-09. Apart from the financial crisis, the Dividend strategy doesn't drop below 20% losses at any time, while the Value model is slightly more volatile.

Testing our models with stop losses

For each of the models I have tested various levels of stop loss between 15% and 45%, both based on the initial opening price and based on a trailing stop. Apart from this, no other changes have been made to the models, so they both:

  • Include all US common shares and depository receipts.
  • Rebalance once a month.
  • Assume a US $7 flat fee commission for each transaction.

If you click through to the InvestorsEdge link above, you can see how each trading strategy works.

Dividend Strategy

Below is a table of the findings from running the Dividend strategy with various stop losses:

Open stop Return Max DD Sharpe
Baseline 18.9% 60% 0.91
15% 16.3% 63% 0.90
25% 18.3% 63% 0.93
35% 17.3% 63% 0.87
45% 16.7% 67% 0.83
Trailing Stop Return Max DD Sharpe
Baseline 18.9% 60% 0.91
15% 10.3% 55% 0.69
25% 15.7% 62% 0.87
35% 18.3% 60% 0.93
45% 17.4% 63% 0.87

You can see from the results of our tests that applying a stop loss of any type and level reduces our returns, sometimes quite dramatically.

From a risk-adjusted return point of view (indicated by the Sharpe Ratio), the best option to select would have been a 25% opening or 35% trailing stop, where you trade slightly lower returns for less volatility.

Value Strategy

Here's a similar table with the effects of introducing a stop loss on our Value strategy:

Open stop Return Max DD Sharpe
Baseline 30.2% 56% 1.10
15% 29.4% 44% 1.15
25% 28.5% 58% 1.09
35% 29.0% 58% 1.09
45% 29.4% 59% 1.08
Trailing Stop Return Max DD Sharpe
Baseline 30.2% 56% 1.10
15% 14.6% 52% 0.84
25% 27.3% 57% 1.12
35% 29.0% 57% 1.12
45% 29.0% 58% 1.08

Apart from the 15% trailing stop, the returns from this strategy aren't nearly as affected as our dividend model. However, as in that model, no stop loss would have improved our bottom line.

What a few stops did achieve was to increase our risk-adjusted returns - an opening stop of 15% was the pick of the bunch, delivering a Sharpe Ratio of 1.15 and minimizing our drawdowns during the 2008 financial crisis to 44% (the S&P dropped 57% during that period).

Your Takeaway

What we have found in analyzing these two models is that, contrary to Investopedia, stop losses don't necessarily make a "world of difference," and when they do, that difference can be dramatically negative.

In conclusion, our two models have shown that they are better suited to stops set against opening prices rather than trailing stops.

Why is that? If we take the majority of strategies on Seeking Alpha, whether they be fully automated ones such as my examples, or more qualitative ones that investors look for stocks with certain characteristics and then do further due diligence before opening positions, they are all typically based on a company's fundamental data. We are looking to buy stocks that are mispriced when compared to a selection of financial metrics. The timing of opening these positions is more an art than a science, and the volatility you experience waiting for a stock price to begin to appreciate can easily be enough to trigger your stop losses.

Our second conclusion is that there is no one stop loss amount that fits all strategies, so if you are going to use them, make sure they either improve your returns or reduce the risk you are taking on.

How do you time exiting positions if you don't trust stop losses? My method is a bit more traditional - if the original reasons for being in a position are no longer valid, or there is a new investment that offers a better return on your capital, then it's time to sell.


You can read comments but to make comments you need to be logged in.

Like what you're seeing?
Try our FREE 30 day trial And test your strategies against historical data immediately