Statistical Arbitrage Between CPA Campaigns and CPM Ad Inventory
Summary
The paper adapts statistical arbitrage to real-time display advertising, where impressions with similar expected effectiveness can carry different prices across fragmented marketplaces or pricing arrangements. Its method links cost-per-action campaigns with cost-per-thousand-impressions inventory. A meta-bidder evaluates each incoming impression bid request against a portfolio of campaigns, using estimated conversion rates, bid landscapes, and other statistics learned from past data to estimate costs and potential profit.
The system combines bidding optimization, aimed at maximizing expected net arbitrage profit, with portfolio risk management that reallocates bid volume and budgets across campaigns to balance risk and return. The authors jointly optimize these components using an expectation-maximization approach. They report support from offline experiments on a large real-world dataset and online A/B tests on a commercial platform across multiple model settings and market environments. The supplied description gives no numerical outcomes or deployment details, so it does not establish how large or durable the gains are. The method concerns advertising auctions, not financial securities markets.
Key ideas
- Fragmented ad marketplaces can create price differences for impressions with comparable expected effectiveness.
- The meta-bidder evaluates CPM inventory against a portfolio of CPA campaigns using historical estimates.
- The method combines expected-profit bidding optimization with campaign-level budget and volume allocation.
- An expectation-maximization process jointly optimizes bidding and risk management.
- The authors report offline dataset experiments and online A/B tests without giving numerical results in the supplied text.
Tags
Full text
# Statistical Arbitrage Mining for Display Advertising # Statistical Arbitrage Mining for Display Advertising We study and formulate arbitrage in display advertising. Real-Time Bidding (RTB) mimics stock spot exchanges and utilises computers to algorithmically buy display ads per impression via a real-time auction. Despite the new automation, the ad markets are still informationally inefficient due to the heavily fragmented marketplaces. Two display impressions with similar or identical effectiveness (e.g., measured by conversion or click-through rates for a targeted audience) may sell for quite different prices at different market segments or pricing schemes. In this paper, we propose a novel data mining paradigm called Statistical Arbitrage Mining (SAM) focusing on mining and exploiting price discrepancies between two pricing schemes. In essence, our SAMer is a meta-bidder that hedges advertisers' risk between CPA (cost per action)-based campaigns and CPM (cost per mille impressions)-based ad inventories; it statistically assesses the potential profit and cost for an incoming CPM bid request against a portfolio of CPA campaigns based on the estimated conversion rate, bid landscape and other statistics learned from historical data. In SAM, (i) functional optimisation is utilised to seek for optimal bidding to maximise the expected arbitrage net profit, and (ii) a portfolio-based risk management solution is leveraged to reallocate bid volume and budget across the set of campaigns to make a risk and return trade-off. We propose to jointly optimise both components in an EM fashion with high efficiency to help the meta-bidder successfully catch the transient statistical arbitrage opportunities in RTB. Both the offline experiments on a real-world large-scale dataset and online A/B tests on a commercial platform demonstrate the effectiveness of our proposed solution in exploiting arbitrage in various model settings and market environments.
Shown in full with attribution under the source's licence. Licence: abstract CC0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.