Larry Lessig argues that four modes regulate behavior in cyberspace: laws, markets, norms, and architecture. How can these four modes regulate the production and spread of fake news? In this paper, we focus on markets and empirically evaluate one particular market-based solution: top ad firms blacklisting fake news producers to eliminate their revenue sources. Our study reveals that fake and low-quality publishers demonstrate a higher tendency to serve more ads and to partner with risky ad servers than traditional news media with similar popularity and age. However, fake news publishers are still strongly reliant on credible ad servers. In fact, the top-10 credible ad servers alone account for 66.7% and 55.6% of fake and low-quality ad traffic respectively. Furthermore, our back-of-the-envelope calculation shows that, at the time of our data collection, the top-10 ad firms were receiving $985.7K to $1.15M monthly from web traffic on fake news sites, a negligible fraction of these firms' annual revenue. Overall, our findings suggest that having top ad firms blacklist known fake and low-quality publishers is a low-cost way to combat fake news.