Why does technical analysis work?

Introduction

If you're here on TradingView, it's probably because you believe that charts and technical analysis can give you an edge in the trading of currencies, metals, cryptocurrencies, and stocks. Granted, sometimes technical analysis doesn't work, but it works often enough to keep hundreds thousands of traders coming back here day after day. The larger question is why.

Four Reasons Technical Analysis (Sometimes) Works

To a fundamental trader like me, technical analysis can sometimes seem like voodoo. Why should lines on a chart tell me anything useful about the total value of future dividends and cash flow for a stock? I admit I especially roll my eyes at Fibonacci ratios. Personally, I feel they're about as scientific as using divination or horoscopes to buy and sell stocks.

But then again, if a lot of people believed that their horoscopes could help them win at stocks, you'd be a fool to ignore them. In fact, you could then gain a large edge by using astronomical data to forecast future horoscopes, getting tomorrow's horoscopes today. Which brings us to the first and most basic reason that technical analysis works:

• It works because people believe it works. If a lot of traders believe that Fibonacci ratios apply to stock markets, then a lot of traders will set their buy and sell orders at significant Fibonacci retracement levels. And then there's another whole contingent of traders who don't believe in Fibonacci numbers, but they know that lots of other people do, so they set their buy and sell orders there anyway. It becomes a self-fulfilling prophecy. Active trading is largely about predicting what other traders will do, and technical analysis is their playbook. And predicting other people's behavior brings us to the second reason that technical analysis works:
  
  
• It works because human psychology follows patterns. For instance, trend-following strategies might work, in part, because of "bandwagoning" and the "Fear of Missing Out" (FOMO). If traders see their friends getting rich off of Tesla or Bitcoin, they will fear being left behind. Speculative enthusiasm cascades through social networks until it has saturated them and everyone is leveraged long to the gills. Only when there's no one left to convert does the momentum finally stall. (Wall Street traders often quip that when their barber starts giving them stock tips, the market is saturated and it's time to sell.) As for support and resistance levels, they work partly because of regret. People remember the price they paid, or the price they wish they had paid, and that memory then shapes their behavior. For instance, if traders remember that they missed several opportunities in 2020 to buy an SPY dip to $323, then they are more likely to buy that level in the event of a future dip. What about oscillators? Well, perhaps humans distrust anything that moves too fast. Even if I'm romantically interested in someone, I'll still pull back if she proposes marriage on the first date. Plus, humans are loss-averse, so at some point we like to lock in gains.
  
  
• It works because it takes time for the market to fully price in news. The advent of algorithmic trading has made it hard for traders to gain an edge by reacting to news events. Stock prices move fast the moment a headline hits, so by the time you see it, you may already be too late. That said, algorithms are pretty good at picking the direction a news event should move a stock, but not necessarily the magnitude. The initial fast news response is often followed by a slow news response as the information spreads through the human population and its implications are assessed and priced by human traders. Trend-following strategies may be able to pick up on these slower processes of repricing in light of news.
  
  
• It works because today's news begets tomorrow's news. This is probably the most underappreciated of all the reasons that technical analysis works. Good news often leads to more good news. If a company posts a large positive earnings surprise, then there's also a good chance that it will get a dividend raise, analyst upgrades, or upward revisions of future estimates in the days or weeks to come. Likewise, bad news often leads to more bad news. For instance, if the company posts a negative earnings surprise, then there's an increased chance that it will need to take on debt or issue shares to sustain operations in the future. The same principle applies to industry-wide or even economy-wide news. If, for instance, the state California bans a company's product, then there's an increased chance that other states will follow suit. And if the Federal Reserve cuts or raises rates, then the next rate change is likely to be in the same direction, because Fed policy goes in cycles. The news-begets-news principle means that trend-following strategies might work, in part, because they are detecting the current direction of the news cascade.
  

Three Reasons Technical Analysis Sometimes Doesn't Work

I should emphasize, however, that technical analysis doesn't always work! Here are a few reasons it might not work sometimes:

• Traders try to anticipate signals. The larger the number of people who know about a trading technique, the less well it works. Take supports and resistances, for instance. If I expect the rest of the market to buy at a particular Fibonacci or moving average level, then I might place my own buy order just above that level in an attempt to front-run everyone else's move. If enough people do this, then the price may not ever actually reach that level.
  
  
• Whales create fake signals in order to harvest profits from technical traders. For instance, if a whale knows that a lot of people have stop loss orders set at a particular support level, then the whale might short a stock to that level in order to trigger all those sell orders, causing a price collapse and an opportunity for the whale to buy shares at a cheaper price.
  
  
• Timing risk. Sometimes you can correctly identify the direction of the trend but still have bad timing. For instance, we're in an interest rate-cutting cycle by the Federal Reserve, which has caused a strong upward trend. But the reality is that we're probably near the end of that cycle. If the Federal Reserve suddenly changed its tune tomorrow and started forecasting rate hikes next year, it would take some time for that information to be fully reflected in slow-moving technical signals, and you could lose a lot of money if you sell only after those signals change. It's perhaps best, then, to have a good understanding of what's driving a technical trend so that you can get out early if you see the underlying drivers change.
  

Comment:

Conclusion

Personally I am a believer in technical analysis. But I also think it's at its best when used with a good understanding of its limitations. For instance, I find that bullish signals work more often than bearish signals, because the market has an upward bias. Thus, I rarely play stocks on the short side, and I almost never use stop losses. I don't need stop losses because, in part, I combine technical analysis with a good understanding of the fundamentals. Even if a key support level breaks, I will still have confidence in the underlying value of my investment. I also like to have a good understanding of the news that's driving technical momentum. If I don't understand why a trend exists, I generally don't play it. Understanding the trend helps me anticipate when the trend might change.