BP bet $8B on a Russian joint venture — then the oligarch partners squeezed it out

BP entered Russia in 1997 by buying 10% of Sidanko. In 2003 it merged its Russian assets with TNK International in a 50/50 joint venture called TNK-BP, creating Russia's third-largest oil producer. The partners on the other side were the AAR consortium — Mikhail Fridman, Viktor Vekselberg, and Len Blavatnik — Russian billionaires who had built TNK through the 1990s privatisations. BP contributed $8 billion in assets. The JV structure gave each side equal board representation and equal veto power, a design that worked only as long as both sides agreed.

By 2007 TNK-BP was producing 1.69 million barrels a day and contributed 24% of BP's global production and 19% of its reserves. But the relationship with AAR was fraying. BP appointed Robert Dudley as CEO. AAR accused BP of running the JV as a subsidiary, hiring costly expats, and shutting Russians out of strategy. In May 2008 AAR publicly demanded Dudley's removal. What followed was an orchestrated campaign: the FSB raided TNK-BP's Moscow office, tax inspectors launched investigations, and the migration service refused to renew Dudley's work visa.

In July 2008 Dudley fled Russia, citing harassment and intolerable conditions. A Russian court banned him for two years. The dispute was resolved in September 2008 with Fridman as interim CEO and restored partner oversight, but the 50-50 deadlock remained. In 2011 BP tried to bypass its partners entirely by striking a $16 billion Arctic share swap with Rosneft. AAR blocked it via the Stockholm arbitration court. The deal collapsed. Rosneft turned to ExxonMobil instead.

By 2012 the partners agreed to sell. Rosneft acquired TNK-BP in March 2013 for $55 billion. BP received $12.3 billion in cash and a 19.75% stake in Rosneft, exiting the joint venture after a decade. The case is taught in business schools as a cautionary tale about JV governance in volatile jurisdictions, the risks of equal-shareholder structures without dispute resolution mechanisms, and the cost of underestimating local partner dynamics.