Deal Or No Deal?

Deal Or No Deal?

Markets are enthralled on Monday with a recycled headline about an imminent US-China trade deal that ran Sunday evening in the Wall Street Journal and was subsequently picked up and expounded upon by Bloomberg and other outlets.

This is just a manifestation of the market’s insatiable appetite for incremental information in the absence of any actual “news”. Late last week, Bloomberg reported that US officials are on the verge of finishing a 150-page “final” trade agreement to be presented at a future meeting between Trump and Xi in Mar-a-Lago, but, as rehashed in our week ahead preview, there are conflicting signals from the administration and the timing is uncertain. Here’s what we said Sunday afternoon:

Steve Mnuchin and Kudlow have variously played “good cop” to Bob Lighthizer’s “bad cop” while Trump has played his traditional role as “insane cop”, tossing out unpredictable soundbites at random intervals. On Friday evening, seemingly out of the blue, he declared that he had “asked China to immediately remove all Tariffs on agricultural products (including beef, pork, etc.) based on the fact that we are moving along nicely with Trade discussions.”

According to the Journal (and subsequent reports), Chinese officials are demanding the removal of tariffs on $200 billion in Chinese goods that went into effect on September 24. Lifting those duties would be a complete 180 from Trump’s threat to more than double them this month. It would also mean un-firing what many analysts described as the first real shot in the trade war. Prior to the implementation of those duties in late September, there was an argument to be made that Trump’s “greatest” tariffs were more bark than bite. It’s also worth remembering that while many market participants trace the Q4 selloff to Powell’s “long way from neutral comments” on October 3, if you annotate your charts to show that comment and the institution of the tariffs on $200 billion in Chinese goods just a week previous, your annotations will overlap. Point being: you could argue that the tariffs were the spark and Powell’s comments were just gas on the fire.

The Trump administration reportedly wants to slow-walk the removal of tariffs in order to make sure China is living up to its promises, while Beijing wants the duties lifted immediately.

As an amusing aside, Bloomberg says “the US asked the Chinese not to retaliate or bring World Trade Organization cases in response to US tariffs that could be imposed to enforce the deal.” That seems like a laughable proposition. Trump has effectively launched an illegal trade war against China and now, as part of the negotiations to resolve the issue, he wants Beijing to agree not to challenge the legality of tariffs he might theoretically slap on in the event he believes the other side isn’t complying with the deal struck to resolve his possibly illegal aggression.

In their piece, the Journal also said Beijing is willing to lower tariffs on US farm products as well as duties on autos and chemicals. China is also reportedly committed to accelerating the removal of limitations on foreign ownership in the auto sector.

Meanwhile, Goldman was out Sunday with a 17-page note aimed at fleshing out what a Sino-US pact “might look like”. Generally speaking, it’s an amalgamation of a bunch of previous notes, but since CNBC has decided to make it the front-page story on their homepage Monday, I suppose it’s worth highlighting a couple of passages and visuals.

After recapping what’s happened over the past couple of weeks, Goldman assigns 75% subjective odds to Trump and Xi announcing a deal contingent upon them meeting in the first place.

“President Trump and President Xi are expected to meet in late March, where they are likely to address issues that remain unresolved after earlier negotiations”, the bank writes. “Assuming this meeting takes place, we believe there is a good chance (75% probability) that the two Presidents will announce a formal agreement of some kind.”

Of course that’s like saying that if I go to the grocery store, there is a “good chance” that I’m going to walk out of there with goods “of some kind”. The chances that Xi is going to fly all the way to Florida, meet with Trump and then leave without making any further progress following recent negotiations are obviously low and the chances of Trump walking away from of a summit with his Chinese counterpart without announcing something big have likely dropped considerably in light of the failed talks with Kim in Hanoi. Trump probably feels like he cannot afford another high profile flop in the space of a month.

In any event, Goldman’s note on this was already dated as soon as it was published. The time stamp on it was 5:56 PM and the Bloomberg headline announcing the Wall Street Journal article crossed around 2:30 on Sunday afternoon. Still, this excerpt from the note was largely in line with media reports:

It is not yet clear whether tariffs will be lifted or at least reduced as part of whatever agreement Presidents Xi and Trump might reach. We would expect that the US would push to keep the current tariffs in effect in the near-term and reduce them only once China implements aspects of the agreement. We expect China to press for their immediate removal. Our base case is that some of the tariffs will remain in place into 2020, but the decision ultimately rests in the hands of Presidents Trump and Xi.

Next, Goldman reiterates the obvious, which is that very much contrary to what Trump shouts at rallies, it is the US consumer and US businesses that pay for the tariffs. That, in turn, means that “only a fraction” of the burden is shouldered by Chinese producers. That reality reduces China’s incentive to placate the Trump administration, but as we’ve seen over the past nine months, the cost to China of letting this farce drag on goes well beyond the mechanical impact of the tariffs themselves.

As far as what China may buy from the US as part of a deal to reduce the deficit (and you’re reminded that, as Janet Yellen put it last week, “almost any economist would tell you that there’s no real meaning to bilateral trade deficits”), Goldman recaps recent media coverage.

“Chinese officials have reportedly contemplated increasing agricultural purchases by $30bn/year, or roughly triple the current level”, the bank writes, adding that China is also “reported to be contemplating increased purchases of crude oil and LNG [and] could target an increase in semiconductor imports from the US of up to $200bn over six years.”

Goldman then endeavors to estimate what China would buy from the US by dividing exports into five categories:

  1. US export and China import discrepancies. For which products is China purchasing relatively less from the US, compared to the share of global imports that are US exports?
  2. Import growth. In which categories has China increased its imports most over the last few years?
  3. Politically strategic US exports. Which US exports are sourced on average from states that support President Trump?
  4. Product complexity. Which US exports are simple, and more easily substitutable, versus very complex, and likely more difficult to substitute?
  5. Exposure to Chinese state-owned enterprises. Do any US exports supply industries dominated by the state in China, suggesting relatively easier implementation of a purchase agreement?

On the first point, Goldman finds that the discrepancy between semis and various other electronics is sizable, as is the gap for oil.



On the second point, Goldman observes that when it comes to what China is importing more of already, energy, semis, and chemicals top the list.

As far as politically-charged US exports, the bank reminds you that “the main agricultural and energy commodities that have been mentioned as under consideration for purchase commitments in the US-China talks also come from states that supported President Trump most disproportionately in 2016.” This is intuitive. Here’s a visual:



On product complexity, Goldman writes that because “complex products are likely to be more differentiated and thus more difficult to substitute between suppliers, China is more likely to increase imports of simpler products from the US.” That’s likely to go further towards explaining why things like agricultural and energy commodities are at the top of the list.

Finally – and this is pretty notable – Goldman writes that when it comes to implementation, the more dominant SOEs are in a given industry, the easier it is for China to implement government-mandated purchases.

“For example, paper products and crude oil both fall into the ‘very simple’ category we identified above, but any purchase agreements of crude oil are probably easier for China to implement, as the crude oil industry in China is dominated by SOEs”, Goldman writes.

This is somewhat ironic – it effectively means that implementing a trade deal predicated in part on allowing for foreign competition in domestic markets will be easier to implement in industries where state-owned enterprises hold sway.

In any event, Goldman rolls all of that up as follows:

Taking these considerations together, we rank 20 goods sectors as shown in Exhibit 8, with the average ranking shown in the far right column (we assign a value of 20 where a category would not have ranked in the top 20). The upshot is that oil and other energy products rank highly across every category; agricultural commodities and semiconductors, which might also be part of China’s offer, also rank highly.


After detailing the extent to which any deal could also involve services, Goldman moves to quantify how much a potential purchase agreement with China could boost US exports at the aggregate level.

The answer is of course “not much”, because even if this hasn’t occurred to Trump, boosting exports isn’t as easy as simply demanding that the second largest economy in the world buy more stuff. Here’s Goldman to explain why:

As noted above, we expect that China would focus much of its total purchase commitment on agricultural and energy commodities, where products from different countries are less differentiated and much easier to substitute among suppliers. In theory, this suggests that increased Chinese purchases of US commodities would be offset with reduced purchases from other trading partners, which would increase ex-US global supply and therefore reduce ex-China demand for US exports of those goods.

That is common sense, something Trump lacks, which is why he doubtlessly hasn’t even considered that this could all be largely for naught. Goldman sums things up as follows:

For example, the US target of $200bn in additional exports over two years, if achieved, would equal an increase in exports to China of 0.4-0.5% of GDP per year. However, net exports would likely increase by much less than this, as the increase in exports to China is largely offset by reduced exports to other trading partners. How much less would depend on the details of the agreement, but we would expect the overall effect on net exports to be a fraction of the headline amount.

So, there you go – some context for any forthcoming deal that sees Beijing promise to boost purchases from the US in an effort to placate a president who steadfastly refuses to listen to anyone other than sycophants and China hawks when it comes to trade strategy.

As far as markets are concerned, it’s just another day, another rumor until someone makes it “official”. Even then, you can expect this to be virtually impossible to enforce, even in the medium term. And if you don’t believe us, just ask Bob.


2 thoughts on “Deal Or No Deal?

  1. Mr. H., thanks for another great piece!
    The most concise take on this topic I’ve read so far.
    I know one’s intuition cannot (and should not!) always be trusted, but glad to see something adressed which I frankly could not wrap my head around so far: Namely that any increase in purchases of whatsover by China had to be offset by reductions in some other areas.
    After all, there is not much ADDITIONAL demand which can be created out of thin air.

Speak your mind

This site uses Akismet to reduce spam. Learn how your comment data is processed.