simple strategy earned 19.79% annualized in 15 years

[roughlyright]

I found this strategy very interesting. Why constantly expose myself to all these risks and waste time, when I can get these sort of results? I am thinking there must be a catch. It cannot be this simple.

 

http://www.valuewalk.com/2016/02/a-simple-strategy-for-income-generation/?utm_source=mailchimp&utm_medium=email&utm_campaign=EMAIL_DAILY&utm_content=quick_link&utm_source=ValueWalk+Newsletter&utm_campaign=9f228f0fd0-%3DUTMDaily&utm_medium=email&utm_term=0_299e40291b-9f228f0fd0-50398245&mc_cid=9f228f0fd0&mc_eid=5eaf686997

[RadMan24]

No. Cherry picking results.

[constructive]

It's a momentum strategy masquerading as a dividend growth strategy. The data set is small in length of time and especially in number of stocks held. They don't provide a credible explanation for why it should work.

[tng]

The problem with just using a bunch of rules is that the returns are almost always driven by something that you didn't originally anticipate. For example, you could end up taking huge sector bets. The good results are almost always a set of rules that is a proxy for avoiding tech stocks during the tech bubble and avoiding finance stocks during the financial crisis. Also, with such a small number of stocks in the portfolio (10 in this case), it is probably one or two stocks that drive the bulk of the returns. So if you keep fiddling with a bunch of rules, you will eventually get the right sector bias that avoids the crises and you end up holding that one company that drives all your returns.

 

You should already be suspect of this strategy going forward because it says to switch to 100% treasuries whenever EPS estimates for the SP500 trends downward. We know the last three decades has been the biggest bond bull market ever as interest rates got cut down to zero. You are not going to get 19.79% returns going forward by shifting into bonds because rates are at zero already. You are not going to get capital appreciation there.

[frommi]

I like the system, its typical quant stuff. This is a value screen (P/FCF<10) with a quality check (dividend+growth) and momentum. And the author added a timing strategy on top. (If forward eps goes down switch to bonds). The only part that i don`t like is the timing part, because this looks like curve fitted to 2000-2002 and 2008. And i wouldn`t use momentum with dividend stocks, i doubt that that really adds a lot of performance. But the author probably tested this a lot before publishing.

In the end the system doesn`t matter, you have to have the eggs to stick to it through thick and thin. This is nearly impossible for 90% of humans especially if you haven`t developed and tested the system yourself.

[ni-co]

The problem with just using a bunch of rules is that the returns are almost always driven by something that you didn't originally anticipate. For example, you could end up taking huge sector bets. The good results are almost always a set of rules that is a proxy for avoiding tech stocks during the tech bubble and avoiding finance stocks during the financial crisis. Also, with such a small number of stocks in the portfolio (10 in this case), it is probably one or two stocks that drive the bulk of the returns. So if you keep fiddling with a bunch of rules, you will eventually get the right sector bias that avoids the crises and you end up holding that one company that drives all your returns.

 

You should already be suspect of this strategy going forward because it says to switch to 100% treasuries whenever EPS estimates for the SP500 trends downward. We know the last three decades has been the biggest bond bull market ever as interest rates got cut down to zero. You are not going to get 19.79% returns going forward by shifting into bonds because rates are at zero already. You are not going to get capital appreciation there.

 

+1

 

Basing anything on dividend yield is a red flag to me because it's so arbitrary.

[roughlyright]

Thank you all for your inputs. What I don't see in the article is what are exact stocks chosen, when. If that information is available, I could verify if the results are true. But I cannot find much information online.