Chinese bank's alternative to Marxism

Is ‘trading economics’ China’s new ‘ideologically acceptable’ theory of capitalism? Michael Roberts looks at the thoughts of a major figure in the People’s Bank of China

In my view, the Chinese economy remains at a structural crossroads.1 The state and state enterprises continue to dominate the economy in investment, employment and production. That means that foreign capital, domestic private capital and market forces do not hold sway, even though they have been increasing in weight and power over the last 30 years.

My view is controversial in Marxist circles. The vast majority of Marxist economists and ‘experts’ on Marx’s ‘theory of the state’ reckon that China is capitalist or ‘state capitalist’. But for me the class nature of the Chinese state remains open.

All I would add at this point is to remind readers of the data that I have previously published on the sheer weight of the public sector and public assets in the Chinese economy.2 The International Monetary Fund has published a full data series on the size of public sector investment and its growth going back 50 years for every country in the world.4 It shows that China has a stock of public-sector assets worth 150% of annual gross domestic product. Only Japan has anything like that amount, at 130%. Every other major capitalist economy has less than 50% of GDP in public assets. Every year, China’s public investment to GDP is around 16%, compared to 3%-4% in the US and UK.

There is nearly three times as much stock of public productive assets to private capitalist sector assets in China. In the US and UK, public assets are less than 50% of private assets. Even in ‘mixed economy’ India or Japan, the ratio of public to private assets is no more than 75%. This shows that in China public ownership in the means of production is dominant - unlike any other major economy. But the IMF data also show that, while public-sector assets in China are still nearly twice the size of capitalist-sector assets, the gap is closing. Private (capitalist) investment stock is growing faster than state-sector assets.


In this article, I want to show that, because the Chinese economy is balanced between the power of the state and the market, this is increasingly reflected in the ideology and economic thinking of Chinese officials and academics. There are still many academics in Chinese universities that hold to what they think is Marxist economic theory and categories. But there are many more - particularly officials in government and state enterprises, who have been educated in ‘western’ universities - who have long abandoned a Marxist view and opted for mainstream neoclassical or Keynesian theory.

A recent striking example is Wang Zhenying, director-general of the research and statistics department at the People’s Bank of China’s Shanghai head office and vice-chairman of the Shanghai Financial Studies Association. This leading Chinese state banker recently summarised his economic views from his Chinese language textbook on economics (for Chinese students) in the Financial Times of all places.

Wang tells us that  “Crises destroy, but also create.” He means that

... the outbreak of each crisis gives rise to new economic theories. Marx’s theory of surplus value was created amidst frequent economic crises in the late 19th century and Keynes’s revolutionary theory was put forward during the great depression in the 1930s. Today, with a worldwide financial tsunami only now receding, people are expecting a new economic theory in response to the failure of the pre-crisis mainstream.5

So for Wang Marx has had his day in the theoretical limelight (ie, 19th century) and, for that matter, so has Keynes (20th century). The recent global financial crisis needs a new theory for the 21st century. Marx and Keynes apparently have nothing more to offer.

And what is this new exciting theory that Wang is proposing to his students to explain the world economic crisis? He calls it “trading economics”:

‘Trading economics’ is one new theory emerging against this backdrop. Mainstream economics deduces the macro whole by extrapolating from the behaviour of individual ‘representative agents’. Trading economics replaces this with a systematic and comprehensive analysis approach. It stresses that, in an interconnected world, the interaction between trading subjects is the fundamental driving force behind the operation, development and evolution of economic systems.

Well, I am still no wiser. Wang explains that mainstream neoclassical theory is stuck with “representative agents” who have “rational expectations” for maximising utility and profits, while market prices move up and down to achieve equilibrium. As Wang says, this bears no relation to the reality of modern economies and never did. In contrast:

Trading economics chooses a different path. Everyone participating in economic activities is put in a specific organisational structure. As a result, their behaviour becomes affected by culture, morality, property and system. There is no ‘economic person’ like Robinson Crusoe in trading economics. Trading economics only has organisations with specific internal structures: households and enterprises. This is the first step to bring economic theory back to the reality.

This all sounds promising. Wang is going to ‘rethink’ economics6 and return it to reality. And what does he come up with? Behavioural economics. And “Behavioural economics experiments have demonstrated that choices are characterised by variety - there is no single answer to all situations.”

Now this is not so promising. Economics is reduced to considering each situation or problem as having a different answer. That would suggest we cannot find any generalised laws about the world economy and its crises.

According to Wang,

Trading economics differs by recognising that different people have different information, in part because they have different experiences. So, while each trading subject seeks maximum profits, the ‘maximum’ differs from one subject to another, even when trading and constraints are the same. Therefore, if the behaviour model in neoclassical economics is the absolute maximum, then it is the relative maximum in trading economics. This is where the difference lies.

Hmm. I am still none the wiser.

What Wang really seems to be arguing for is free trade and international integration:

From the study of the development rules of economic system, it is found that global economic integration is neither the innovation of a single politician, nor a strategy implemented by a certain country to pursue its own interests. Instead, it is the only option following the development rules of social economic system. Today’s world has shifted from an isolated island, through small-world development, to a network without marks. To achieve comprehensive progress and development, the world economy must promote the integration of trading network among countries. We need to listen to the warning of trading economics in a world awash with anti-globalisation thoughts.

This is shades of the very line presented by president Xi Jinping at Davos 2017, where he claimed that China is the leading globaliser. Now the economic theorist of the People’s Bank of China is offering “trading economics” to support Xi.

A key feature of Wang’s “trading economics” is that it rejects the idea of looking for causal relationships between economic variables. He refers to the “masterpiece” work of rightwing monetarist Milton Friedman’s analysis of the cause of the great depression of the 1930s. Friedman argued that the failure of the Federal Reserve Bank to properly control the money supply was the cause. The banks collapsed because of an unnecessary monetary squeeze. But others argued that the economy collapsed because of a change in ‘expectations’.

Wang concludes: “It is impossible to find a single factor among various events to explain the great contraction.” You see, it is just too complex for ordinary mainstream theory. So Wang says we must “give up on simple causal relationships”. Instead, using “trading economics”, we can get “a concrete structure through the trading network.” Then, apparently, “various possibilities of economic operation can be predicted, including the fluctuation of economic cycles, the probability of crisis, assessment of policy effects, etc”.

Wang provides no evidence in his FT article for his claim of the power of prediction enabled by trading economics. And here is the nub of his theory: namely its close “connection between macroeconomics and behavioural economics”. According to Wang,

The behaviour of each trading subject and the ways in which they react to external disturbances can be informed by the research of behavioural economists and psychologists. The economic operation simulated in this way is better targeted and the analysis has more solid experimental foundation.

‘Great leap’

There we have it. Far from carrying out empirical research for cause and effect, all we need is to go back into the laboratory of behaviour and do ‘experimental research’. Wang claims that this “is a great leap in methods of economic theory research, because it represents the unity of economic research and natural science in methodology”.

Actually, a behaviourist approach is an economics cul ­de ­sac. Before the global financial collapse, this micro-motivation approach to economics was popular with young economists, who had turned away from questions like poverty, inequality or unemployment to study behaviour on television game shows. Looking at the ‘irrational’ behaviour of people’s brains and thinking was substituted for the aggregate trends and changes in modern economies.

The irony of Wang’s view is that, since the global financial crash, empirical studies have come back into favour in looking for the causes of the great recession, because mainstream and behavioural theory had failed. Despite that, Wang wants us to ditch Marxist macro theory for Keynes’s psychological ‘animal spirits’ or the micro ‘nudge’ theories of behaviourists like Richard Thaler.7

Is the way forward really through behaviourists developing computer models, where the idea is to populate virtual markets with artificially intelligent agents who trade and interact and compete with each other, much like real people? Sure, every situation is different, but anyone who makes a living out of data analysis knows that ‘heterogeneity’ is limited enough, so that the well ­understood past can be informative about the future.

In my view, if economists want to understand the causes of financial and economic crises, they need to look away from individual behaviour models and instead look to the aggregate: from the particular to the general. And they need to turn back from deductive a priori reasoning alone towards history: the evidence of the past. History may not be a guide to the future, but speculation without history is even less based in reality. Economists need theories that can be tested by evidence - but the evidence of the aggregate and history, not the laboratory.

Yes, Wang recognises that mainstream economics is no good at explaining developments in modern capitalism, but does “trading economics” take us any further? It seems more like an ideologically acceptable theory as an alternative to Marxism in the country of ‘socialism with Chinese characteristics’.

Michael Roberts blogs at


1. See



4. November 6 2017:

5. See

6. See