How to Be a "Connoisseur of Extremes"

[MadMarketScientist 11]

From The Daily Crux

 

The Daily Crux: You've mentioned in the past that to enjoy a lifetime of trading success, you've got to be able to spot "extremes" in the market… that you must become a "connoisseur of extremes."

 

What do you mean by that?

 

Brian Hunt: By saying you should become a "connoisseur of extremes," I'm saying you should always be searching for situations where a market is in a drastically different state than normal.

 

By locating these extreme states – and then betting on conditions returning in the direction of normal – you can consistently make low-risk profits in any type of market.

 

It's important to realize that extremes can occur in any market – from stocks to commodities to real estate to bonds to currencies.

 

Extremes can be fundamental in nature… like how cheap or how expensive a stock market is. Another name for this is a "valuation" extreme. Extremes can also be price-action-based… like how overbought or oversold a market is. That's a "technical" extreme. And extremes can show up in sentiment readings, like surveys that monitor investor pessimism and optimism.

 

Crux: Let's cover valuation extremes…

 

Hunt: Sure. A good example of a fundamental valuation extreme came in U.S. stocks in 1982. Back then, stocks became extremely cheap relative to their earnings power.

 

For U.S. stocks, the normal price-to-earnings multiple over the past hundred years or so is 16. In 1982, the economy and the stock market had been doing so poorly for so long that people simply gave up on stocks. Since nobody wanted to own stocks, they became extremely cheap. The price-to-earnings multiple fell to around 8.

 

It was one of the greatest times ever to buy U.S. stocks. The market rose 50% in just one year. It doubled by 1986. It rose more than 10-fold over the next 17 years.

 

Fast-forward about two decades, and you find the opposite extreme. In 1999, optimism toward stocks was so high that the market reached a price-to-earnings ratio of 33. This was a ridiculous, extreme level of overvaluation.

 

Remember, the normal price-to-earnings ratio of the past 100 years is around 16. The extreme level of overvaluation made it a terrible time to buy stocks. The market crashed for several years after hitting that extreme.

 

When it comes to fundamentals, you need to study an asset's historical valuation and find out what's normal for that asset. When an asset gets very cheap relative to its historical valuation, you need to consider buying. When an asset gets extremely expensive relative to its historical valuation, you want to consider avoiding it… or even betting on it falling.

 

This goes for oil stocks, tech stocks, real estate, and lots of other assets.

 

Crux: OK… so people need to buy stocks when they get extremely cheap relative to their historical norm and avoid them when they get extremely expensive relative to their historical norm. How about extremes that are "technical" in nature?

 

Hunt: Before we get into particulars, let's define the term to prevent confusion.

 

Technical analysis is the study of price action and trading volume. Many people think technical analysis is all about predicting the market, but it's not. It simply comes down to using price and volume data to gauge market action… and to help guide decisions. That's it.

 

There are dozens of technical indicators that measure a stock's oversold/overbought levels. One I've found useful is the "RSI," which stands for "relative strength index." The RSI is nothing magical or predictive. It's simply an objective way to gauge the overbought/oversold nature of a stock.

 

My colleague Jeff Clark is amazing at finding short-term technical extremes in the market. He uses an indicator called the "bullish percent index" to identify overbought/oversold extremes in broad market sectors. I'm sure Jeff will tell you there's nothing magical or predictive about the bullish percent index. Again, it's simply an objective way to gauge price action.

 

We are using these gauges to identify extremes in the market… then betting on the conditions being "relieved" in the other direction. When the pressure behind an extreme is released, the market tends to snap back like a rubber band stretched to its limit.

 

There are literally hundreds of technical indicators and chart patterns people use. While I have a handful of things that I know work, what works for me or you or someone else isn't as important as knowing the overarching goal: That you're using this stuff to spot extremes and trade them.

 

For example, I often trade short-term moves in blue-chip stocks, like Coke and McDonald's. These are elite businesses with tremendous competitive advantages and long histories of treating shareholders well.

 

But like any business, even stable blue chips go through rough patches. If they report a weak quarter or have a product recall, or any other of a dozen solvable problems, the market tends to overreact and sell the shares. The stock price will reach a state we can term "oversold." This is a condition where the stock has reached an extreme level of poor short-term price action.

 

It's around this time that I'll step in and trade the stock from the long side. World-class businesses have a way of rebounding from short-term setbacks. They tend to snap back from extremely oversold levels.

 

Crux: OK, when it comes to technical analysis, we're looking for extreme conditions that when relieved, produce "snap back" moves.

 

You mentioned extremes in sentiment. Let's cover that idea…

 

Hunt: Let's also define this term to prevent confusion. The study of market sentiment comes down to gauging the amount of pessimism or optimism toward a given asset. You can gauge the sentiment for just about any kind of asset… be it stocks, commodities, real estate, or currencies.

 

Gauging market sentiment is more art than science. There are lots of ways to gauge sentiment that cannot precisely be measured… and some that can.

 

Whatever gauges you use, the goal is the same: to find extreme levels of pessimism or optimism. You want to find situations where the majority of market participants are extremely bullish or bearish… and then bet against them. You want to go against the crowd.

 

When most folks can't stand the thought of owning a particular kind of investment, chances are good that it's cheap… and that it's due for at least a short-term rebound.

 

On the other hand, when everyone loves an asset – like when everyone loved stocks in 1999 – chances are good that the asset is expensive and due for at least a short-term drop.

 

A few informal sentiment gauges – the kind that can't be precisely measured – are magazine covers and cocktail party chatter.

 

If a mainstream publication like Newsweek or Time has an asset on its cover, chances are good that the asset is far too popular, far too expensive, and due for at least a short-term drop.

 

Magazine publishers have to write stories lots of people want to read. Plus, it's mostly journalists – not great investors – who write those stories. Mainstream magazines are just going to write about what's popular so they can sell lots of magazines. Back in 1999 and 2000, they always had stocks on their covers. It was a danger sign. In 2006, it was all about how to cash in on the real estate boom. That was a danger sign.

 

The idea behind studying cocktail party chatter is similar. It's another way to get a feel for what the general public thinks about a given investment.

 

You can get a feel for this by talking to people at cocktail parties, family gatherings, holiday parties, and dinner parties. When lots of people are excited about a given asset and are buying as much as they can, it's a major warning sign. It's a sign the asset is too popular, too expensive, and due for a fall.

 

On the other hand, when most folks can't stand the thought of owning a given asset, chances are good that it's a good buy.

 

For example, back in 2003, I put a large portion of my net worth in gold. When I'd tell people that I owned a lot of gold, they'd look at me like I was crazy. You could say there was an extreme amount of disinterest in gold. Gold went on to rise many hundreds of percent.

 

Crux: What are some sentiment indicators that can be measured precisely?

 

Hunt: Money managers and investment newsletter writers are always being surveyed and monitored.

 

Just like most regular investors, the supposed professional investors get swept up in crowd-following behavior. You want to bet against extremes here as well.

 

Crux: It sounds like being a "connoisseur of extremes" is all about finding abnormal situations, and then betting on them becoming normal again.

 

Hunt: Exactly. It's important to note that being a "connoisseur of extremes" – and trading them – is about getting a powerful force of nature to work in your favor. That force is called "reversion to the mean."

 

"Reversion to the mean" is a broad term that is used to describe the tendency for things in extreme or abnormal states to return to more normal states. You see "reversion to the mean" all the time. You see it in academics, business, trading, and dozens of other areas.

 

For example, winning an NFL Super Bowl requires an extreme set of circumstances. A football team has to have a great coach… a great set of players… and they have to play extremely well for an extended period of time. Its elite players have to avoid injury. It has to beat a series of excellent teams at the end of the season.

 

It's really hard to get all the stars aligned and pull off a Super Bowl-winning season. That's why Super Bowl winners tend not to win the championship the next year. They tend to "revert to the mean" and not win it.

 

To go back to the example of trading extremely oversold blue-chip stocks, if a blue-chip stock like Coca-Cola is sold heavily day after day for several weeks, chances are good that its trading action will "revert to the mean" and cease being so extreme. Chances are good that it will stop falling and start rising.

 

Crux: Understood. Any final thoughts?

 

Hunt: One last thing that I think is important to note is that an extreme in valuation is often accompanied by extreme technical and sentiment readings.

 

That's why I believe studying and trading the market with "just" fundamentals or "just" technicals can be a limiting mindset. Consider what happened with offshore-drilling stocks in mid-2010, just after the terrible Gulf of Mexico oil well disaster.

 

After the disaster, investors dumped shares of offshore-drilling stocks. They completely overreacted. It was like people believed we'd never be drilling for oil again. Sentiment toward the sector was terrible. Even companies with little business exposure to the Gulf of Mexico fell more than 30%.

 

This big decline left the whole sector in an extremely oversold state. It also made the stocks very cheap. Great drilling businesses were sold down to valuations of around five times earnings.

 

After the selloff, you had a sector that was extremely unpopular, extremely cheap, and extremely oversold from a technical standpoint. So, I went long offshore drilling stocks and made big returns in a short amount of time.

 

The stocks enjoyed a sharp "snapback" rally. Again, this rally was preceded by "extreme" valuation, technical, and sentiment readings.

 

Crux: That's why it pays to look for extremes of all types.

 

Hunt: Yes, exactly.

 

Crux: Thanks for your time.

 

Hunt: My pleasure.

[Enigmatics 10]

Haven't even read this yet, but looking forward to it.

 

Just from the title alone I have to say I firmly agree. Conceptually I would rather take reversion to the mean (VWAP) or mode (POC) trades than mess around anymore with everything in the middle, breakouts, breakdowns, etc. etc. There is just so much nonsense (algos, backfilling, etc) that goes on in those particularls areas and so much more assumed downside risk that people don't realize.

 

I've seen various market statistics suggesting markets are ranging 70-75% of the time so why not use it to our advantage?

 

I'm finding that it's become a situation where I have to force myself not to trade anymore unless I get those reversion opportunities. That has actually become the hardest part of all of this. As a good peer of mine said to me earlier today:

 

"It takes a ton of discipline to accept the fact that the market, despite directional bias and macroeconomic events, will simply backtest supply/demand to proceed in any given trajectory - people want to make that Babe Ruth homerun call, not simply navigate the market back inside to a prior days value area, but at this juncture it's simply the best r/r set-up available and I don't see anyone really talking about it nonetheless taking advantage of it."

Edited April 3, 2013 by Enigmatics

[SunTrader 10]

Brian Hunt forgot to mention 1980 and 1981 was sideways leading into the 1982 bull market start point so PE's were probably low for those years as well.

 

Finding an extreme is only half the battle. Guessing right timing-wise is the other half.

[SIUYA 11]

it all makes sense - its just that actually defining what the mean is, how extreme is extreme and then how you either work out you are wrong, or work out how to re-enter again becomes much the same issue as any strategy.

Plus too many mean reversion players still have to cut things and not fall into the trap of blowing up when the mean moves towards and surpasses their entry!

 

Anyone who went through the internet bubble might be less inclined to talk about mean reversion.....even when combining the fundamentals and technicals.

Like everything it has its place and Enigmatics sums up the issue regardless of style....its hard to wait, while Suntrader is also right....guessing the timing is the other hard part.....

 

another circular never ending discussion on its way

[DbPhoenix 42]

it all makes sense - its just that actually defining what the mean is, how extreme is extreme and then how you either work out you are wrong, or work out how to re-enter again becomes much the same issue as any strategy.

 

Plus too many mean reversion players still have to cut things and not fall into the trap of blowing up when the mean moves towards and surpasses their entry!

 

Anyone who went through the internet bubble might be less inclined to talk about mean reversion.....even when combining the fundamentals and technicals.

 

Like everything it has its place and Enigmatics sums up the issue regardless of style....its hard to wait, while Suntrader is also right....guessing the timing is the other hard part.....

 

another circular never ending discussion on its way

 

This is essentially what Auction Market Theory is all about. What "mean reversion" players often don't understand is that mean is dynamic. There is a mean even in a trend. The extremes are determined in large part by the inability of traders to find trades.

[Enigmatics 10]

it all makes sense - its just that actually defining what the mean is, how extreme is extreme and then how you either work out you are wrong, or work out how to re-enter again becomes much the same issue as any strategy.

Plus too many mean reversion players still have to cut things and not fall into the trap of blowing up when the mean moves towards and surpasses their entry!

 

Anyone who went through the internet bubble might be less inclined to talk about mean reversion.....even when combining the fundamentals and technicals.

Like everything it has its place and Enigmatics sums up the issue regardless of style....its hard to wait, while Suntrader is also right....guessing the timing is the other hard part.....

 

another circular never ending discussion on its way

 

See I don't consider it guess work anymore. I've been using divergence, volume profile, and volume spread analysis to assess the points in time during an intraday trading session when supply or demand have been maxed out. But again the hardest part is waiting for that to show itself. It's too easy sometimes as traders to fall in love with what visually looks like a "possible" directional move, especially early in the session when things are very liquid. I've been in quite a few of those in my time and have just grown tired of the door slamming shut as the instrument suddenly decides "nope, we're gonna range instead today."

 

I look at it this way. Even if the instrument opens up bullishly, how do I really know what the target is on the intraday level? Traders all use different points of reference as targets (i.e. Fibs, Pivots, etc.). How do I know how much volume is going to come in via traders or on-open market orders? With mean reversion, the concept is a little more cut and dry. Supply or demand dry up and the stock will revert back to the mean or mode to get people interested again. Anyone who observes the behavior of instruments in regards to their VWAP or POC can see this almost every single day, barring a parbolic trend day which rarely ever happens.

Edited April 3, 2013 by Enigmatics

[SIUYA 11]

Enigmatics - fair enough....but what you are discussing in terms of range days is certainly not a point of extremes as per the OP conversation.

Hunt: "I'm saying you should always be searching for situations where a market is in a drastically different state than normal"

 

hence if the normal type of day is a range bound one, then Hunt is likely looking for extremes of a non range trading day, or a series of non range trading (or normal) days.

 

It sounds like you must have determined that either the day is likely to be range bound and is not a range day, OR is at least not likely to accelerate too far from your entry.....those rare parabolic days.

The question for me would become - if the points of entry on a range trading day become some clear, then are they just as clear on a trending day when the mean, vwap, poc etc; is moving and you can go with that trend......

 

As for the guess work, if you have clearly defined rules etc then its not guess work, but I would be skeptical if any one told me they had a sure fire way of picking tops and bottoms of extreme movements - (and yes they can coincide with a massive support or resistance etc....but this is a deviation between intraday and day trading as well.)

 

I know I know - hence the never ending circular discussions and this is not to show right/wrong/truth in anything.....but simply to me - anyone who can pick good turning points is still better off trading with a trend than against it whilst too many think that mean reversion is simply a matter of trading against a trend. (if that makes sense)

[Enigmatics 10]

Enigmatics - fair enough....but what you are discussing in terms of range days is certainly not a point of extremes as per the OP conversation.

Hunt: "I'm saying you should always be searching for situations where a market is in a drastically different state than normal"

 

hence if the normal type of day is a range bound one, then Hunt is likely looking for extremes of a non range trading day, or a series of non range trading (or normal) days.

 

It sounds like you must have determined that either the day is likely to be range bound and is not a range day, OR is at least not likely to accelerate too far from your entry.....those rare parabolic days.

The question for me would become - if the points of entry on a range trading day become some clear, then are they just as clear on a trending day when the mean, vwap, poc etc; is moving and you can go with that trend......

 

As for the guess work, if you have clearly defined rules etc then its not guess work, but I would be skeptical if any one told me they had a sure fire way of picking tops and bottoms of extreme movements - (and yes they can coincide with a massive support or resistance etc....but this is a deviation between intraday and day trading as well.)

 

I know I know - hence the never ending circular discussions and this is not to show right/wrong/truth in anything.....but simply to me - anyone who can pick good turning points is still better off trading with a trend than against it whilst too many think that mean reversion is simply a matter of trading against a trend. (if that makes sense)

 

Nothing is surefire. That's why it always comes down to money management and risk tolerance.

 

When it comes to intraday reversion to the mean trading it boils down to the probabilities. What is more likely to occur? The stock parabolically moving up the entire day? Or will it end up ranging? Statistics show the former by a wide margin.

Edited April 3, 2013 by Enigmatics

[SIUYA 11]

Engimatics - if the normal days are a range and you trade mean reversion....then how would you define an extreme day or how do you then determine when to apply the reversion to the mean on extreme days.

 

do you do the same process?

[Enigmatics 10]

Engimatics - if the normal days are a range and you trade mean reversion....then how would you define an extreme day or how do you then determine when to apply the reversion to the mean on extreme days.

 

do you do the same process?

 

Again, I've been predominantly using divergence and volume analysis. Chances are though that on an extreme trend day I will not attempt a reversion trade. Typically the volume profile doesn't lend to it, nor does the VWAP ....which commonly has a higher, tighter slope. That usually doesn't leave a desireable trade because there's not much "vig" between where supply/demand ran out and VWAP and that's what I'm trying to capture.