Via Westpac’s excellent Robert Rennie:
Back in November we introduced a simple iron ore port inventory model (specifically an inventory deviation model – see “Negative iron ore supply shocks waning“) that could be used as a useful indicator to point to where iron ore prices might settle down to if we saw signs that the incredibly positive Chinese demand shock and modestly supportive negative Brazilian supply shock were showing signs of normalising.
Now while we are seeing more signs that the negative supply shock is waning, with a noticeable pickup in combined Brazilian and Australian export numbers in January and positive momentum continuing into February, we are yet to see any real concrete signs that Chinese demand is actually slowing.
Indeed, while November and December Chinese iron ore import volumes did slow, shipping activity data suggests very strong imports for the month of January. So, we would caution against arguing that prices are headed back towards our ‘equilibrium’ target near term given evidence of ongoing Chinese demand.
However, the purpose of this note is to add a little bit more detail to the iron ore inventory model that we published back in November following some discussion with industry experts.
The additions we have made fall into two buckets. Firstly, on top of using the deviation of the current level of port inventory from its 5yr seasonal average, we have added a second deviation measure which captures the deviation of high quality ore inventory (Brazil and Australian) from its 5yr seasonal average.
The second change we have made reflects a desire to get a better sense of the relationship between inventory and demand. To do so, we have added two additional models. The first of these expresses port inventory as days of imports; the second as days of steel production.
If we then take these four models, plot an average of the forecasts and throw standard deviation bands reflecting the forecast error, we get a single model that points to the current iron ore equilibrium being in the region of $105 to $130. That is, given the availability of supply at port, importers should be willing to pay $105 to $130, rather than the recent extreme highs, assuming the last 5yrs of price history is a guide.
Now as before, we would caution that using inventory models with a 5yr sample period introduces several potential risks. At times, inventory can be divorced from supply and demand, especially if exogenous shocks are in play. Arguably the enormous increase in Chinese demand over the last year is one such example of an exogenous shock.
The other concern that we would highlight is the fact that we have not seen prices above $130 for iron ore over the rolling 5yr model period – bar the recent break higher. This is why current prices look so out of line with historical 5yr price relationships given that inventory is not stretched.
There is also some evidence that part of the increase in inventory may be intentional. We start 2021 with the inventory of ore from Brazil at record levels both for this time of year and historically. Given that La Nina conditions tend to be associated with higher rainfall in Brazil, which can impair export volumes, some of this port inventory rise could have been precautionary.
The final caveat is that VLOC shipping activity from Brazil was on the rise through 2020H2 as new port capacity opened in China. January saw a strong start to the year for departures, suggesting this story will continue in 2021, giving Chinese ports greater access to Brazilian supply.
With those caveats though we would emphasise that recent break up to $170 looks super-normal given current inventory levels when we adjust for history, quality, iron ore import demand and steel production-based demand.
Thus, we would be surprised if prices pushed back anywhere close to those levels again, at least on the basis of current inventory availability. We maintain the view that prices will head lower once demand and supply dynamics normalise – with a move back to $105 to $130 being our target.
However, the answer to the question ‘when will iron ore fall below $130’ can only be answered with another question – ‘how long will Chinese demand remain at all-time record levels’. Given evidence of China building strategic reserves across an array of hard commodities, only time can answer that question. For the record, Westpac forecasts iron ore below $130 in Q3 this year.
I reckon $130 in May and sub-$100 in September with higher before and after.