Technology grouping and quantitative strategies: When the market is no longer “average”

> This English version was automatically translated from the Chinese original. Please refer to the original Chinese article if any wording differs.

Technology grouping and quantitative strategies: When the market is no longer “average”
Many people understand quantitative strategies and easily think of them as a set of stable operation formulas: stock selection, sorting, buying, position adjustment, and then waiting for long-term probabilities to materialize.
But real markets aren't always this benign.
In some stages, market opportunities are diffuse. Many industries have capital participation, and small ticket, mid-cap, growth, value, and themes take turns to perform. At this time, quantitative strategies are very comfortable because there are enough samples, enough spreads, and enough factor opportunities in the market.
But there are also phases where the market suddenly becomes very concentrated.
Money is no longer evenly distributed, but is flowing in a few directions. For example, artificial intelligence, semiconductors, computing power, chips, science and technology innovation board, robots, innovative drugs, or dividends, special valuations, and resource stocks at certain stages. Indexes may still be rising and popular ETFs may be hitting new highs, but a large number of stocks in the market are falling.
This is the most typical feature of the technology group market: a few sectors have sucked up most of the liquidity.
It looks lively, but it doesn't necessarily represent market-wide health. For many quantitative strategies, this is a more difficult environment.
Quantitative stock picking relies on market breadth.
The so-called market breadth is not just about the number of rising stocks, but more importantly, whether the opportunities are dispersed. If funds are willing to flow among many industries, many market capitalization levels, and many styles, then factor models have room to play. Factors such as low valuation, growth, quality, reversal, momentum, small market capitalization, etc. can all find differences between different stocks.
But in the technology group stage, the market will become very narrow.
The strong direction will continue to be strong, and the weak direction will continue to be drained. A large number of stocks fell not because their fundamentals suddenly turned bad, but because the money was not there for them. At this time, if the quantitative strategy still mechanically selects stocks in the entire market or small-capitalization stocks, it may fall into an embarrassing situation: the stocks selected by the model look cheap, have excellent factors, and have a good historical winning rate, but there is no money to buy them in the short term.
This is not because the model is completely invalid, but because the market environment has changed.
Technology grouping is essentially a change in capital structure.
When funds are concentrated in a few technological directions, individual stock selection becomes difficult. Because within the same popular sector, there are great differences between leaders, followers, concepts, and pseudo-concepts. If you buy the wrong stock, the index may go up, but your account may not; the sector may go up, but your position may not.
At this time, the significance of ETFs becomes prominent.
ETFs do not need to determine which technology stock is the real leader, nor do they need to bear risks such as individual stock announcements, performance, reductions, suspensions, and black swans. As long as a certain direction does become the main line of the market, ETFs can more directly take over the capital trend of this main line.
Therefore, in the group market, ETF is not only a defensive tool, but in many cases it is a more suitable offensive tool.
If a technology ETF, chip ETF, science and technology innovation ETF, or GEM ETF has active transactions, a stable trend, and continues to attract money, it may be more suitable for the quantitative system to follow than most individual stocks. Because the price of ETFs has comprehensively reflected the selection of funds within the sector, the noise is lower than that of a single stock.
This leads to a key question:
Quantitative strategies should not only judge "what to buy", but also "what method is suitable to buy now".
When market opportunities spread, quantitative stock selection of individual stocks has advantages. Because opportunities are widely distributed, the model can screen out targets with better odds from a large number of stocks.
Sector or thematic ETFs have an advantage when market opportunities are concentrated. Because funds are concentrated and the main line is clear, ETFs can avoid making repeated mistakes in the differentiation of individual stocks.
Therefore, technology grouping and quantification are not contradictory relationships. The real question is whether quantitative strategies can identify whether the current market is a "diffused market" or a "concentrated market."
A more natural way of observation is to look at three types of signals.
First, look at whether the industry momentum is highly concentrated.
If a few industries continue to rise while most industries weaken, it means that funds are grouping together. Especially when the technology direction continues to occupy a strong position, the main line of the market has become relatively clear.
Second, look at whether market breadth has deteriorated.
If the index does not fall much, but the number of falling stocks increases significantly, or small and medium-sized stocks and micro-cap stocks continue to weaken, it means that the rise has not spread to most stocks. At this time, the environment for individual stock quantitative strategies usually becomes worse.
Third, see if strong ETFs continue to emerge.
If the strong direction not only rises, but also has active trading volume, smooth trend, and small retracement, it means that the funds are not a one-day trip, but are forming a phased main line.
Putting these three signals together is more meaningful than simply looking at the rise and fall of the index.
When the index rises, it does not mean that the market is easy to do; when the index fluctuates, it does not mean that there are no opportunities. What really matters is whether money flows across the market or is concentrated in a few directions.
For quantitative strategies, the most dangerous thing is not a market decline, but a misjudgment of the market structure.
If the market has entered a technology group, and the strategy is still to select stocks based on the general spread of the market, it will be easy to continue to be sucked.
If the market has shifted from grouping to spreading, and the strategy is still all on ETFs, it is possible to miss a large number of individual stock recovery trends.
Therefore, a more reasonable quantitative framework is not to stick to one style forever, but to switch risk exposures under different market structures.
When the market spreads, let individual stock strategies play more roles.
When markets are concentrated, let ETF momentum strategies take on more positions.
When the market is extremely concentrated, you can even turn to ETFs completely to avoid consuming funds in weak stock pools.
When the market picks up again, quantitative positions in individual stocks will be gradually restored.
The idea behind this is not complicated: quantification is not for stubbornness, but for disciplined adaptation to the market.
The technology group market will make many traditional stock selection models feel uncomfortable because it compresses the distribution of opportunities and changes the market from "multiple blooms" to "a few main lines". But it also provides another direction for quantitative strategies: no longer just looking for opportunities from individual stocks, but looking for opportunities from the capital structure.
When the entire market is discussing which technology stock is the strongest, the quantitative system can take a step back and ask a more important question:
Are funds concentrating?
If the answer is yes, then following strong ETFs may be more effective than looking for the perfect answer in individual stocks.
If the answer is no, it means that market opportunities are still scattered, and there is room for individual stock quantitative strategies to continue to play.
Ultimately, technology groups are not the enemy of quantitative strategies. It just reminds us that the market is not always suitable for trading the same way.
A good quantitative strategy should be able to identify both individual stock opportunities and market structure; it should be able to make breadth of money in the diffuse market and mainline money in the group market.
What is truly worth pursuing is not predicting when the next technology grouping will begin, but when it has already happened, the strategy can recognize it, follow it, and exit in time when it ends.