Confusion about the validity of smart money addresses

I became curious about the concept of smart money recently and tried to uncover some smart money addresses and their trading patterns. But the results were disappointing.

A particular token was screened. It had high social media hype in the past month, large market cap, and high trading volume. In short, this token still seemed to be trending upwards with good momentum.

An uptrend was captured during which the needed smart money addresses could be filtered, from the price bottom to the price top. Some tools were utilized, attempting to uncover the smart money addresses through metrics like order size, frequency of trades, timing of trades.

^{Order size was judged based on personal experience, but smart money trade sizes probably wouldn't be too small.}

^{Very high trading frequencies were filtered out, as those are more indicative of bots connecting via API.}

^{For trade timing, orders that sold at price highs were chosen, since buying low and selling high is the simplest effective profit strategy.}

In this way, over 30 addresses were found, and 5 hours were spent backtesting their trading records. But the results weren't ideal. A few factors:

^{Some lower frequency traders had held the token for a while, likely traded once on a whim, profited, and continued holding. The profit didn't sustain, clearly not smart money.}

^{Some high frequency traders poured assets into other tokens after taking some profits. Their assets were still vulnerable to market swings.}

^{Some traders, making only 2-3 trades a month, hold and sell when price is right, sometimes at a loss, sometimes profiting, but win rate decent.}

...

There is uncertainty if smart money is effective, but markets are driven by emotions and narratives. Much smart money likely bought in early, but timing of selling is highly uncertain.

If there are better methods for uncovering factors related to smart money addresses, please feel free to comment! Thank you all.

​