Looking at the actual charts presented in the article, I can't help but draw a very different conclusion: the spikes in Bitcoin-related HN posts do not precede

Looking at the actual charts presented in the article, I can't help but draw a very different conclusion: the spikes in Bitcoin-related HN posts do not precede BTC price increases, but instead they more or less mirror the price swings (as do most other relevant news sources). It's deceiving, especially at the scales used, because it's easy to compare the peak of a price swing with the uptick in news stories (implying some kind of predictive power) when in reality, the upticks in stories are fairly proportional to the price changes leading up to the price peaks - it's just not as noticeable due to the difference in scales. As with other markets, a trading strategy based on public news is - for any normal trader - a losing strategy. The people who are able to make gains based on news in other markets often have to pay a premium to colocate servers at or near the exchange's servers, and a trade delayed by more than a few milliseconds will have lost any potential upside in the market's reaction to the news. You can bet on the same thing happening with Bitcoin (and the decentralized nature of Bitcoin doesn't change the requirement to be located near the exchanges' servers since these trades all happen outside the blockchain - i.e., on traditional servers).