If you’ve read my previous post, you’ll probably be thinking, “here he goes… another stock market bear getting ready to spew bull market hate.” This time, however, I’m going to do my best to show the market in a neutral perspective. In fact, I’m a huge stock market bull…. from 50,000 ft.
Anyone can blindly stare at an S&P 500 chart and say that stock values are at absurdly high levels, therefore everyone should sell before it all comes crashing down. Historically, however, the stock market has always produced a positive annual return over the long run. Maybe those buy and hold preachers aren’t just blowing smoke, maybe they’re really on to something. After all, they have Warren Buffet on their side.
To get a better view of stock market performance over the long run, I like to use the annualized rolling average. The charts below show what your annual return would have been over various lengths of time, on a rolling basis. Meaning, if I purchased one share of the S&P 500 in 1900 (yes, I realize it didn’t exist then), what would it have been worth in 1905? How about 1901-1906, or 1902-1907. If you carry these calculations out all the way to current, you’ll have the 5 year rolling average of the market. Then do it again for 10, 20, and 30 years. That’s a whole lot of calculations.
Anyone can blindly stare at an S&P 500 chart and say that stock values are at absurdly high levels, therefore everyone should sell before it all comes crashing down. Historically, however, the stock market has always produced a positive annual return over the long run. Maybe those buy and hold preachers aren’t just blowing smoke, maybe they’re really on to something. After all, they have Warren Buffet on their side.
To get a better view of stock market performance over the long run, I like to use the annualized rolling average. The charts below show what your annual return would have been over various lengths of time, on a rolling basis. Meaning, if I purchased one share of the S&P 500 in 1900 (yes, I realize it didn’t exist then), what would it have been worth in 1905? How about 1901-1906, or 1902-1907. If you carry these calculations out all the way to current, you’ll have the 5 year rolling average of the market. Then do it again for 10, 20, and 30 years. That’s a whole lot of calculations.
You’ll notice that the volatility of the returns goes up significantly as the time frame shortens. The 5 year rolling average swings from over 30% annually to -13%, while the 30 year chart only fluctuates between 2% and 11% (see chart below). More importantly, the 30 year chart never has a negative time frame. The minimum was around 2% in 1932, meaning you had purchased in 1902 and just before the 30 years ended, the Great Depression hit, wiping out the majority of 30 years’ worth of gains, but still squeezing out 2% annually. The inflation adjusted average return (horizontal red lines) calculated to 6.75% for the 30 year average.
Historically, buy-and-hold really has worked over the long run. Through World War I, the Great Depression, World War II, the Korean War, the Vietnam War, the Civil Rights Movement, two oil shocks, the Dot-com bust and the Financial Crisis, the 30 year rolling return still averaged just under 7% during the last 116 years.
Does anything in the rolling average charts lead you to believe that the stock market is going to correct in the near future?
I could take 10 different investors, and they’d probably interpret the charts in different way. For example, the 10 year chart could be interpreted to say that the prices will keep rising, based on the shape of the previous movements and the fact that the market doesn’t tend to stick around the mean line very long.
Does anything in the rolling average charts lead you to believe that the stock market is going to correct in the near future?
I could take 10 different investors, and they’d probably interpret the charts in different way. For example, the 10 year chart could be interpreted to say that the prices will keep rising, based on the shape of the previous movements and the fact that the market doesn’t tend to stick around the mean line very long.
The 5 year chart shows a declining trend that could be alluding to a decrease in prices.
The more an investor tries to forecast short term movements based on any chart, I think they’re setting themselves up for failure. Nothing in our markets is short term anymore. However, with that said, the data is definitely encouraging that over the long run, a prudent steadfast investor that doesn’t pull their money out when times are tough will end up with positive returns.
The real questions is: How many investors could stay invested all the way through a market crash?
These returns are only applicable for an investor that never pulled their money out when times were tough, and while it’s easy to say you personally wouldn’t succumb to the fear and sell after a crash, many investors end up throwing in the towel and selling anyway. The Buy-High and Sell-Low motto will always run strong during a correction, and it’s the worst possible move an investor could make during such a time.
While buy-and-hold is great for probably the majority of investors. For those of us that like to study the markets and choose individual stocks, we’re always looking for ways to improve our methods, knowing full well that if we can’t beat the market averages over the long run, then we’re just doing ourselves and our accounts a disservice.
The easiest method is controlling the purchase price, but as you’ll see later, that alone isn’t enough to guarantee better returns then the average. Just as you wouldn’t run into Walmart and buy a bed after seeing a “prices have increased 30%!” advertisement, so you shouldn’t with the stock market. The current Shiller PE multiple is my preferred method of gauging how expensive the market is relative to its long term average.
The real questions is: How many investors could stay invested all the way through a market crash?
These returns are only applicable for an investor that never pulled their money out when times were tough, and while it’s easy to say you personally wouldn’t succumb to the fear and sell after a crash, many investors end up throwing in the towel and selling anyway. The Buy-High and Sell-Low motto will always run strong during a correction, and it’s the worst possible move an investor could make during such a time.
While buy-and-hold is great for probably the majority of investors. For those of us that like to study the markets and choose individual stocks, we’re always looking for ways to improve our methods, knowing full well that if we can’t beat the market averages over the long run, then we’re just doing ourselves and our accounts a disservice.
The easiest method is controlling the purchase price, but as you’ll see later, that alone isn’t enough to guarantee better returns then the average. Just as you wouldn’t run into Walmart and buy a bed after seeing a “prices have increased 30%!” advertisement, so you shouldn’t with the stock market. The current Shiller PE multiple is my preferred method of gauging how expensive the market is relative to its long term average.
It’s intuitively obvious that if you pay more up front, your future returns will probably be lower than someone who paid less up front. The 30 year rolling returns vs starting price are detailed in the plot below and generally confirm this sentiment. Historically, the average Shiller PE is 16.69. The current number is 26.6 (mid Sept 2016). There is very little data at that high of a PE, therefore it's hard to estimate what kind of future returns those purchasing into the market will get. My guess is that it'll be at least a couple percentage points lower than the long term average.
Although the rolling returns do have a negative slope, they’re also fairly noisy. At an initial purchase Shiller PE of 20, the rolling return ranges from 3.3% to ~8% annually. That’s a difference of over $70,000 on a $10,000 initial investment. In my opinion, other than for the extreme low and high PE multiples, the evidence isn’t solid enough to anticipate an expected long term portfolio return based solely on the purchasing PE multiple.
What if I sell when the PE multiple gets high, and buy back in when it gets low?
I hear this statement quite often, and decided to build multiple models to see if I could buy and sell strictly based on the PE multiple, and subsequently beat the buy-and-hold average. A few models are detailed below and they all start with the same assumption. You buy one share of the S&P 500 in 1900 (based on data compiled by Shiller), and you reinvest all dividends.
Model 1: Sell when the PE reaches 50% over the long term average (i.e. 25 on an avg. of 16.69), and buy back it when it drops to 80% of the long term average (13.3)
What if I sell when the PE multiple gets high, and buy back in when it gets low?
I hear this statement quite often, and decided to build multiple models to see if I could buy and sell strictly based on the PE multiple, and subsequently beat the buy-and-hold average. A few models are detailed below and they all start with the same assumption. You buy one share of the S&P 500 in 1900 (based on data compiled by Shiller), and you reinvest all dividends.
Model 1: Sell when the PE reaches 50% over the long term average (i.e. 25 on an avg. of 16.69), and buy back it when it drops to 80% of the long term average (13.3)
This method slightly underperformed the buy-and-hold strategy. It was especially bad in the 20 year time frames ending from about 2009 on, mainly because the Shiller PE elevated high enough to sell out, then never retreated to the 80% mark to purchase back in. Not a good model to base your retirement on.
Model 2: Sell when the PE reaches 50% over the trailing 5 yr average, and buy in when it drops to 75% of the trailing 5 yr average.
Model 2: Sell when the PE reaches 50% over the trailing 5 yr average, and buy in when it drops to 75% of the trailing 5 yr average.
This model was looking to capitalize on the PE trend. If it’s rising, the sell point would rise with it. It would only sell if the current PE spiked significantly above the average. Similar to the first model, there were various 20 year spans in which this method provided a higher return. However, by the mid 1990’s the PE had increased so fast, it caused the model to sell out and the future gains were lost. The overall rolling average was worse than Model 1. Another bad portfolio idea if the twenty year time frames were ending between 1992 and 2008.
Rather than just focusing on the PE multiple, lets base buy and sell decisions on current earnings as compared against trailing averages.
Model 3: Sell when the current earnings drop to less than 75% of the 1.5 year average earnings. Buy back in when the earnings are 25% above the 1.5 yr average. It’s essentially trying to catch the short term earnings cycles.
Rather than just focusing on the PE multiple, lets base buy and sell decisions on current earnings as compared against trailing averages.
Model 3: Sell when the current earnings drop to less than 75% of the 1.5 year average earnings. Buy back in when the earnings are 25% above the 1.5 yr average. It’s essentially trying to catch the short term earnings cycles.
This method actually looks pretty good for the 2000 to current time frame. Although the total rolling average is slightly lower (6.2%), it looks to be a better approach than the previous models. This model handled the big market swings between 1997-Current better by only selling when the earnings topped out, which was subsequently near the market peaks.
Model 4: Recognizing that Model 1 did better in the early years, and Model 3 did better recently, this model combined the parameters of the two models, upped the level of complexity, and attempted to realize the good parts of both models. Sell when the earnings drops to less than 75% and the PE reaches 50% above the long-term average. Buy back in when the earnings are 25% above the trailing average.
Model 4: Recognizing that Model 1 did better in the early years, and Model 3 did better recently, this model combined the parameters of the two models, upped the level of complexity, and attempted to realize the good parts of both models. Sell when the earnings drops to less than 75% and the PE reaches 50% above the long-term average. Buy back in when the earnings are 25% above the trailing average.
This model looks very similar to the buy-and-hold, with a slightly better result in the recent years. Definitely the best of the 4 simple models.
While I only showed 4 example models here, I tried over 10 different versions combining earnings and PE multiples. In every case, they underperformed the buy-and-hold strategy. These are obviously simplified models. However, I often hear investors talking about buying and selling based on PE values, or strictly earnings, with no regard to surrounding factors such as interest rate levels, GDP growth, foreign market stability, etc. Therefore, I wanted to explore how these methods would have done historically.
One big takeaway I noticed in the looking at the raw data was the importance of the reinvested dividends. In any of the models, selling at high PE valued did help capture the peaks in terms of market price, but knowing when to get back in, and losing all of those dividends in the mean-time more than negated the benefit of selling high in the first place. In addition, none of the models took into account brokerage fees or taxes, which would have made all of the results worse. Model 4 would have also underperformed if fees and taxes were incorporated.
Putting my money where my mouth is, if you read my last post I did mention that I’m transitioning to cash. However, being in cash doesn’t mean I’m not doing anything. As I said above, as soon as I sell I’m losing dividends. Therefore I had to make up that money in some way, and I’ve been using cash-secured puts to generate the supplemental income until the prices come down to what I feel are reasonable levels.
In the end, it’s no surprise why most highly regarded professional investors stress that many investors would be better off with a buy-and-hold strategy. If you truly have the time, and more importantly the interest to research global economies, dive into company financials, and do the work necessary to truly understand what is happening within a specific company or the world, then you could consistently beat the market. But let’s not fool ourselves, most of us don’t, and won’t. Therefore, buying and holding an S&P 500 index fund is probably answer for the majority of the population. Just don’t let Wall Street hear you say that.
While I only showed 4 example models here, I tried over 10 different versions combining earnings and PE multiples. In every case, they underperformed the buy-and-hold strategy. These are obviously simplified models. However, I often hear investors talking about buying and selling based on PE values, or strictly earnings, with no regard to surrounding factors such as interest rate levels, GDP growth, foreign market stability, etc. Therefore, I wanted to explore how these methods would have done historically.
One big takeaway I noticed in the looking at the raw data was the importance of the reinvested dividends. In any of the models, selling at high PE valued did help capture the peaks in terms of market price, but knowing when to get back in, and losing all of those dividends in the mean-time more than negated the benefit of selling high in the first place. In addition, none of the models took into account brokerage fees or taxes, which would have made all of the results worse. Model 4 would have also underperformed if fees and taxes were incorporated.
Putting my money where my mouth is, if you read my last post I did mention that I’m transitioning to cash. However, being in cash doesn’t mean I’m not doing anything. As I said above, as soon as I sell I’m losing dividends. Therefore I had to make up that money in some way, and I’ve been using cash-secured puts to generate the supplemental income until the prices come down to what I feel are reasonable levels.
In the end, it’s no surprise why most highly regarded professional investors stress that many investors would be better off with a buy-and-hold strategy. If you truly have the time, and more importantly the interest to research global economies, dive into company financials, and do the work necessary to truly understand what is happening within a specific company or the world, then you could consistently beat the market. But let’s not fool ourselves, most of us don’t, and won’t. Therefore, buying and holding an S&P 500 index fund is probably answer for the majority of the population. Just don’t let Wall Street hear you say that.
RSS Feed
