IIF: EM ex China outflows since the war began are not unprecedented

• EM ex China outflows since the war began are not unprecedented.

• In contrast, outflows from China in Q1 were the largest on record, and were broad based, affecting equities, FX bonds, and local govies.

• Local bond sales by foreigners were particularly large in Feb-Mar, and unlikely to be fully explained by Russia’s use of its RMB assets.

• Investors may be reassessing exposure to high geopolitical risk EMs.

The impact of war in Ukraine on EM portfolio flows has not been catastrophic so far.

Our daily trackers registered outflows from EM ex China, but the episode does not come close to the most severe in the last decade.

For example, fast increasing long-term US bond yields early last year triggered larger outflows.

The situation in China is radically different. Our portfolio flows tracker, which maps well to lagged official data, points to the largest quarterly outflows on record.

Foreigners sold equities and bonds, with localcurrency bonds accounting for most of the outflows.

Capital flight stands in sharp contrast to the large inflows in 2020-21, when investors increased their exposure to Chinese bonds by 30-40% each year.

Reserve sales by Russia may explain some of the outflows at the beginning of the war but we are unconvinced they are driving this episode.

We think a combination of covid lockdowns, depreciation, and perceived risk of investing in countries whose relationships with the West are complicated explain capital outflows from China.

As a corollary, the renminbi still seems far from a top global reserve asset.

Our suite of high-frequency portfolio flows trackers is useful to assess the impact of events such as war in Ukraine in real time. Unlike other high-frequency flows products, our data map well to official balance of payments data that come out with a long lag.

The picture from our series captures the full universe of portfolio investors. In this note, we focus on flows to China, which have been unusually weak since Russia invaded Ukraine. Even though China recorded inflows in January, outflows in February and March were large enough to make 22Q1 the worst quarter on record (Exhibit 1).

Equities saw outflows that continued into April, but the relevant action was in bonds (Exhibit 2).

Outflows affected both FX and local-currency bonds (Exhibit 3). The former probably reflect concerns around real estate developers. The latter were the largest and the ones where the global geopolitical background matters the most. Outflows from government and development bank bonds started in February and picked up in March (Exhibit 4). Taken in isolation, outflows in Q1 are not outsized relative to other countries.

Outflows from government bonds in 22Q1 add up to 1% of the position nonresident investors held in December 2021.

The equivalent figures for India and Indonesia are 4.1% and 4.7%. What makes China’s bond flows remarkable is their sharp swing negative.

Investors increased their allocation to China significantly through the covid crisis, in part due to bond index inclusion in 2020 (Exhibit 5). Some argue Russia selling part of its roughly $70bn in reserves allocated to China explains outflows.It is a point worth considering, as we estimate a non-negligible part of flows to Chinese govies in the last few years were reserve accumulation by central banks (Exhibit 6).

Russia may have used RMB assets in the second half of February, when total reserves declined and access to dollar and euro assets became complicated. It is less clear that in March sales by Russia were large as reserves were stable. If anything, part of the current account surplus may have ended in Chinese bonds, which are free of sanctions.