Session 31 (of 42): Market Timing - Non-financial and technical indicators

0:00 Hi. Welcome back. In this, my second session on market timing, I want to talk about some of the indicators people use to try to time markets. I will look at non-financial indicators and what they are, as well as technical indicators, things we've talked about in the context of charting. So, let's start with non-financial indicators. Non-financial indicators, as I said, don't show up in your Wall Street Journal or maybe in your stock price pages, but I would classify non-financial indicators into three groups. The first are what I would term spurious indicators, things that look like they're correlated with the market, but have no reasonable basis for it.

0:38 Now, we look at what's, for instance, the Super Bowl indicator. What does that mean? Why does it work? You know, if it does work, what does it tell us? The feel-good indicator measures how happy investors are feeling. Presumably, happy investors are more likely to push up prices. I'm not sure whether that's a good sign or a bad sign, but many people actually feel that if people are feeling too good, they get hyped up. That's a sign of a market bubble. There's going to be a correction coming. So, let's start with spurious indicators.

1:08 There are dozens and dozens of indicators that people claim tell you what's going to happen in the market. Now, when you look at the statistics backing it up, most of the time the research is, "This is beyond the reach of statistics." And standing alone, that is actually true. For a long time, for instance, we talk about the Super Bowl indicators. Those of you don't know how the Super Bowl works, the Super Bowl is an American football phenomenon, where the winner of There are two divisions in American football, the NFC and the AFC.

1:42 And the winner of each division meets at the end of the Super Bowl, one of the most, talked about, watched phenomena in sports, perhaps only second to the World Cup. And one of the things researchers noticed, at least until very recently, was if that Super Bowl match-up was won by the AFC team. AFC and NFC are the two divisions. Markets are more likely to go up. So, as an early example of how researchers presented this, here's how it go.

2:12 There've been 35 Super Bowls played in and in 28 of those 35, this indicator worked at predicting what the market would do. In other words, AFC or the NFC. Now, 28 out of 35 is too high to be pure chance, and technically that is true. That beats the 50/50, which you'd expect with randomness. But, here's the problem. If you have hundreds and hundreds of things out there, indicators that you have data on, and you try all all of them, some of them will beat the odds of chance, even if they're purely driven by luck.

2:49 Second, some of these spurious indicators give you a sense of market direction. For instance, you'll you'll hear that if happens, markets are more likely to go up than down. Remember, up and down is a pretty weak prediction. A market going up 5% might not be a good year. It's so many of these indicators don't look at do these markets go up more than expected or less than expected. And finally, if you find an indicator where there's no economic rationale. And again, I hate to go back to the Super Bowl indicator, but I'll go back to it.

3:22 There is no economic story you can tell for why there should be a link between who wins that game and what happens to stock prices. So, spurious indicators are indicators that look good on paper, but almost never work in practice, partly because they're entirely driven by chance. Let's move on to feel-good indicators. Let's start with a statement that I think is true. When people feel upbeat, they feel optimistic, they push up stock prices because that optimism feeds in there, but it also affects the rest of what they do, the styles and the social mode.

3:57 The 1920s, for instance, good time for markets, of course, was also the period where Americans partied. They felt good. 1950s, the same way. So, over time, it's not surprising that people have found a link between social indicators and Wall Street. Now, I saw one that claimed that if you know how many bankers are eating at high-priced restaurants in New York City, you can tell me what the market is doing, and that's true. There is a high correlation between demand for high-priced food and services in New York City and how well the market is doing because much of the demand is coming from people who work in financial markets, and when markets are doing well, they're more likely to be doing well as well.

4:38 But, here's the catch. That's a contemporaneous correlation. You're saying, "What does that even mean?" Well, when markets are doing well, people party more, right? But, knowing that people are partying more is not helping you on markets because your job as an investor is to get ahead of the game. You need a lagging indicator the markets to be lagging your indicators so you can observe the indicator and say, "That's what's going to happen to markets next year." You'll see that as a common theme in much of market timing. Knowing that something is correlated with markets is not useful if it's contemporaneous. It happens the same time. You need to get ahead of the game to be able to time markets based on that indicator.

5:20 Third, there are hype indicators. The essence of a bubble is people are hyped up. They're driving up prices. They're being irrational. And there are arguments to be made that when there is a fad and investors get caught up in the fad, it's going to show up elsewhere. It's a classic indicator used in Wall Street, so all the cocktail party chatter indicator, which is when you go to a cocktail party, how quickly does the talk turn to stocks? In New York, it turns to stocks almost immediately.

5:48 But, let's say in Des Moines, Iowa, and 2 minutes into a cocktail party, people start talking about stocks. That's an indicator that hype has risen to the surface. A variant of this is how quickly your Uber driver, or whoever your ride-sharing driver is, starts talking about stocks. The argument being that if, you know, and and there's an old John F. Kennedy's father was a legendary investor in Wall Street in presumably got off got out of the market before the correction, before the great, you know, the great correction. And the story is he knew that it was time to get out of the market when the shoeshine boy, at those times, you had people set up shoeshine stations on the street, and people who were going to work would stop and get their shoe When a shoeshine boy started talking about which stocks to buy.

6:38 That's a hype indicator. And as investors turn to social media, what's cocktail party, the Uber driver, the the shoeshine boy is being replaced by what are people saying on Twitter? That the same hype that drives that the old indicators is now showing up in what people tweet. So, maybe by looking at social media, you get a sense of how much hype is driving it. Now, of course, the you know, while there is some basis for hype indicators, deciding what's unusual can be tricky to do. What's an unusually upbeat mood on markets? That I mean, defining abnormal can be tricky when what's normal keeps shifting, when standards and tastes are changing.

7:21 And even if you can observe a hype indicator, that things are getting hyped up, I'm not sure how you follow through. You think that must mean there's there's a correction coming? You might be right, but what if it takes 5 years between the time you observe an indicator and the correction happens? So, hype indicators might work, but there's there might be no timeliness in when they work. So, those are the non-financial indicators. The second group of indicators are indicators we talked about in the context of individual stocks that might be used to time markets, technical indicators.

7:54 Now, as with individual stocks, there's a tension here between the momentum and the reversal players. So, if you remember technical indicators broadly grouped, you have momentum indicators who believe that what's happened in the past is more likely to continue into the future. And reversal indicators who think that if things have been really good in the past, there's going to be a correction coming. The same tension plays out in markets where the same indicator can be bullish for one group and bearish for the other.

8:22 In addition with the market, people find information in trading volume just as they do with the individual stocks. A surging volume or a drop in volume might be a good or a bad sign that you bring into views in the markets. As well as where the market volatility goes up and down, side indicators that you can observe by looking at other instruments in the market. So, broadly speaking, is there information in knowing what the market has done in the last year, the last 2 years in predicting what the market will do in the future? Here's what I did to try to answer that question.

8:56 I have over 153 years of stock prices and I thank Robert Shiller for putting that data going back to 1871 here. I took the 153 years and broke them down into what stocks do after a down year and an up year. Let's focus on the down year and the up year first. There were 36 down years and 36 up years. Just one down year, one up year. Right? The return in the following year was 10.14% after a down year, 12.05% after an up year. Statistically, those are very similar.

9:29 If you have two up years, right? Then, numbers start to shift. It turns out that the next year is more likely to be a bad year. The return is only 1.78%. You say, "There, I have my strategy. If there are two up years, I'm going to sell stocks." Well, the problem is after two down years, your returns are also not really good. So, two down years and two up years are both bad news. Already, you can see without even digging deeper and deeper into this data that the the the more you get into the depths of the data, the more confused you're going to be because the patterns are shifting and there's no clear signal you're getting by looking what markets did in the last year. Now, you can try this by quarter, by month, but one of the things you're going to see is timing markets based purely on what markets have done in the past has not worked that well.

10:23 Now, there's also a variant of that. If you remember when we talked about timing patterns, we talked about the January effect that the Jan- the January historically has been the best month of the year. In fact, all of the small cap premium we talked about came in January. So, usually the way people present this argument is if you know what happened in January, you can predict what the year will look like. And technically, you're right. A bad January usually goes with a bad year for stocks and a good January goes with a good year for stocks. You say, "This is good. I'll wait till the end of January, and if January is a good if I have a good January, stocks are up a lot, then I'm going to invest in stocks. And if stocks are down in January, I will not invest in stocks." Here's the problem.

11:07 While it's true that January returns are correlated with the returns over the overall year, remember by the time you observe January returns, January is already behind you. You're going to invest from February through December. And if you look at those 11 months, the signal becomes much weaker. Put simply, I can't really statistically show you that the returns in the remaining 11 months are going to be any different whether you have a good January or bad January, they kind of even out.

11:36 January indicator looks good at first sight, not so good if you take a deeper look. What about trading volume? Here the story is cut both ways. There's an argument that if you see stocks per stock prices going up but with light trading volume, that's a bearish sign. There's a correction coming. At the same time, there are people who argue that if you see very heavy trading volume, there's a correction coming as well. You think, "That's contradictory." One of the things about technical indicators is they are contradictory. The same indicator can give different signals to different traders depending on their perspective of that indicator. So, trading volume affects prices, yes, but is it Is it a Is the relationship something you can use to time markets? I'm not sure. Maybe you can find it as you start digging through intraday or daily or weekly trading volume.

12:28 There might also be information in the option market. Remember we talked about index futures in the context of arbitrage and index options. Now, remember when people get upbeat about stocks, they might buy call options, a leveraged bet on stocks going up. And if they feel pessimistic, they might buy put options. There are people who observe the rate the trading volume on puts versus the trading volume on calls. So, it becomes a contrarian indicator. When the trading volume in puts picks up relative to calls, people are getting more pessimistic, there are people out there who view that as a sign that the market is going to turn around. Again, data and trading volume either in the market stock market itself or in the option market.

13:11 There are also people who observe money flow. At the risk of stating the obvious, if money is coming into markets, I don't know from where, that's a good sign. Stocks are more likely to go up. So, if you can find indicators of that money flow, both the direction and the amount, it might help you predict markets. So, let's say your indicator tells you that money flow is going to go up that it's going to flow into the market over the next 6 months for whatever reason. You might buy stocks because that money flow is highly correlated with returns. But that correlation is again contemporaneous. In periods when money flows into markets, stocks tend to go up. But for you to make money on money flow, that's not sufficient, right? You got to forecast what money flow is going to look like in the next 6 months in the next year. And if you can do that, there is there is informa- there is at least backing to the notion that you might be able to time markets.

14:06 There There are researchers who extended the study to global equity markets and they find evidence of momentum. That when prices are going up, they're more likely to go up. Prices are going down, they're more likely to go down across geographies. But they find that the evidence is stronger in high trading volume markets and weaker in low trading volume markets. Finally, volatility matters. And until about 40 years ago, we talked about volatility, but it was not observable. You couldn't trade on it.

14:38 But in the mid '80s, you had indices that allowed you to trade not at the level of the market, but on market volatility. The most widely used of these indices is called the VIX, the volatility index. And basically, it's a bet on what the volatility of the market will be. So, that number might So, let's assume the VIX is 25 and that's what the historical number has been. You're around the average VIX. And usually when people draw high, low, or typical numbers, they're drawing it based on history.

15:12 So, this is a study that looked at what happens when the VIX is much higher than expected bas- based on with expectations based on the past and much lower than expected. In the period where the VIX jumps, it increases, it's not good for stocks. You see stock prices go down. Right? In periods where the VIX decreases, stock prices go up. So, in in the contemporaneous period, an increase in VIX goes with lower stock prices. For those of you watch markets, this shouldn't come as a surprise. The days where the market is down are the days where the VIX jumps.

15:48 But, you can't make money on it. It's contemporaneous. But, here's where it gets interesting. When the VIX increases, bad news for stock prices now, but historically, it's been good news in terms of what stocks do in the subsequent period. So, when volatility increases, it turns out that stocks do much better in the subsequent period. When volatility decreases, stocks do not as well in the subsequent period. And that's a tradeable strategy. You look at the VIX, it's higher than expected, you buy stocks. It's lower than expected, you sell stocks.

16:25 Whether you will make enough money to cover transactions costs and taxes, that's going to be a tricky question because these are not big surges in returns, 1 and 1/2%. It's not something that you can just use to cover up large transactions costs, but there is clearly a link between volatility and what you see happening to markets. There are a whole set of other indicators that I won't go through because all of those chart patterns we talked about with the individual stocks can also be used at the market level.

16:54 Remember support and resistance lines? There are people who use support and resistance lines for the entire market, trend lines for the market. Again, the battle between momentum and reversal plays out in these chart patterns. There are also what I call sentiment indicators. What do sentiment indicators try to capture? The fear, greed. They're trying to you know, measure the balance between fear and greed. They're trying to see how fearful are people getting because they're getting fearful stock prices should go down.

17:25 I'll give you a simple example of a sentiment indicator. It's called the confidence index. So, you compare the yield on a triple B rated bond. That's at the very edge of investment grade to a triple A rated bond or to even to a treasury bond. As people get more scared, the spreads increase. Not surprisingly, and when the spreads increase, it turns out that risk premiums in equity markets also go up, which pushes down stock prices, sentiment indicators.

17:55 And finally, there are indicators of what the traders' sentiment is. Are they getting bullish or bearish? And people again use this either as an indicator that, you know, bullishness will carry over as a contrarian indicator. So, remember mutual fund cash positions. We talked about this as a way in which, you know, mutual funds decide stocks are not, you know, the place to be. You You put it into cash. So, when mutual funds get down on markets, they get bearish, they increase their cash positions as a percentage of total hoard. There are people who track mutual fund cash holdings. And if they're contrarians, here's what they do. When mutual funds get more bearish, the cash holding goes up.

18:37 They say They buy stocks because they say mutual funds are getting more bearish, that's a good sign for markets. Same with investment advisers. You know, when they get more bullish, that's viewed as a If you're a contrarian, as a bad sign for markets. So, again, dozens and dozens of other indicators, but the bottom line with all of these indicators again is you got to ask two questions. One is are the indicators leading indicators? Do they tell you what's coming? Are they contemporaneous? You want leading indicators. Second, how strong is the linkage? How noisy is it? Because if you're going to be right only one time out of 10, it might not be enough. And third, will the returns you make by following these indicators cover the transactions costs and taxes that come from trading more frequently?

19:23 So, here's the bottom line on, you know, finan- non-financials and technical indicators. There are dozens, perhaps even hundreds of indicators that are correlated with market movements. And we can see that up front. But, for you to be able to make money on them, first they have to be observable. So, you have a hype indicator where you say, "You got to observe the mood of people." You can't I I you can't go around gauging people's moods. So, you got to use something that I can use to observe the mood.

19:51 Second, they have to lead markets. Again, I hate to sound like a broken record, but to make money on an indicator, it's not enough that the indicator moves with markets, but that observing the indicator tells you what markets will do in the next period. And finally, it's got to be big enough to cover transactions costs and trading costs. Most of the analysts that are Most of the indicators that I see analysts using to time markets tend to be contemporaneous rather than leading and not useful as investment signals.

20:19 So, you need to separate the signals from the noise. And I hope you succeed. I hope you found the session useful, and I thank you very much for listening.