How Harvey And Irma Will Slam The US Economy: A Complete Walk Thru From BofA

Last week, before the full devastation from Hurricane Harvey was unveiled, Goldman and JPM were the first banks to suggest that the storm's impact on US GDP would be modest: a slight decline in Q3 growth,  offset by a similary modest rebound in Q4 and further as emergency funds "trickled down" through the economy. Now, with more clarity on just how destructive the storm has been, other banks are coming out and they are not nearly as confident that the damage from Harvey will be "modest" - in fact, according to a just released analysis from Bank of America, Harvey will result in at least a 0.4% hit to Q3 GDP, which has reduced BofA's Q3 GDP estimate to 2.5%.... and that excludes Irma.

Here's Michelle Meyer explaining why in just a few weeks, all the economic misses will be blamed on, you guessed it, hurricanes.

First came Harvey, next comes Irma


Hurricane Harvey crashed down on the shores of Texas, leaving behind record flooding and destruction. According to early estimates, Harvey may end up being the most expensive natural disaster in the US since 1980, costing $70-108bn (Table of the day). Thousands of people have been impacted. We are now actively monitoring Hurricane Irma, which threatens to hit the coast of Florida over the weekend.


BofA attempts to provide a rough guide of how to quantify the impact on the economy, by exploring the following channels:

  • High frequency data: Harvey has already shown up in initial jobless claims contributing to a 62,000 jump in claims for the week ending 2 September. We also expect some impact in consumer confidence, consumption, construction, industrial production, and inventories.
  • GDP: Natural disasters tend to reallocate growth - serving as a drag in the quarter when the disaster hits and a boost in later periods from rebuilding efforts. We estimate that Harvey will end up slicing 0.4ppt from 3Q GDP tracking, bringing our forecast to 2.5%. Hurricane Irma threatens to drag growth down further. Rebuilding is historically a long process, implying upside risks to growth early next year.
  • Inflation: Energy prices have spiked, but pressures will likely ease. There are upside risks to auto prices given replacement demand.
  • Washington DC: The urgency to pass a bill to provide funds for hurricane relief facilitated a bipartisan compromise for a three-month continuing resolution and suspension in the debt limit.

Harvey - and then Irma - will complicate the interpretation of upcoming data releases. In the short term, the hurricanes serve as a drag, but ultimately history suggests the rebuilding efforts underpin growth. Unless this time is different of course. In any event, the Fed will be flying blind (in a hurricane so to speak) in a few days when it has to decide whether to hike rates further following the latest economic data.

The early signs to watch for Harvey (and Irma's) impact:

The first evidence of Hurricane Harvey's impact on the economy was the larger-than-anticipated drop in August unit auto sales to 16mn SAAR, released last Friday. Initial jobless claims for the week ending 2 September spiked to 298,000-the highest level since April 2015-driven by filings in Texas. We took a look at various high frequency indicators during the past hurricanes, listed in Table of the day to get a sense for how the data might evolve.(*) In particular, we compared the monthly trajectory following the disaster to the six-month average level or growth rate leading up to the event. In other words, how much did the indicator deviate from the prior underlying trend? Since Harvey is shaping up to be one of the costliest hurricanes in history, we calculated a weighted average trajectory across past disasters based on the real cost of damage.


The post-Harvey moves in initial jobless claims and auto sales are thus far consistent with our findings. History also suggests auto sales remain depressed for a few months following the hurricane, which means we should not expect a quick rebound (Chart 1).


BofA's analysis indicates the next signs of Harvey will likely show up in the soft data, namely consumer sentiment and confidence, which also proves to be somewhat persistent. Weakness will also emerge in real consumption, industrial production, business inventories, and construction spending. These will have implications on GDP tracking, as discussed below.

Then there is Inflation, where as anyone following the price of gas at the pump, it's "all about energy"

Energy and autos are two categories in the inflation data most sensitive to natural disasters. Prices at the pump have already jumped higher, increasing 30 cents or almost 13% in the immediate aftermath of the Hurricane. This likely reflects challenges distributing and producing gasoline, as more than 20% of US oil-refining capacity was knocked out as a result of Harvey. However, prices should normalize as refineries get back online over coming weeks, meaning only a transitory bump to headline inflation. Indeed, we have already seen wholesale gasoline futures return to pre-Harvey levels, which will ease pressures at the retail level (Chart 2).


We see some upside risk to auto prices. The prices of autos have been on a persistent downward trajectory given weakening demand and excess inventory. Many autos in Houston will need to be replaced, boosting demand and potentially tempering some of the downward pressure on prices. That said, our auto analysts are skeptical that the replacement demand will be enough to offset the headwinds from the turn in the cycle.


Hurricane Irma also threatens to impact near-term inflation as it could impact $1.2bn of crops in Florida, most of which are fresh produce. Irma could greatly reduce supply of these crops which would lead to higher prices for a period of time. The bottom line is that there is upside risk to headline inflation as a result of the hurricanes and possibly slight upside to core inflation.

Which brings us to the economic punchline: why BofA believes that GDP: weaker now, stronger later

We estimate Hurricane Harvey will slice 0.4ppt from GDP growth in 3Q, which takes our tracking estimate down to 2.5%. Hurricane Irma, if it hits Florida as feared, would also be a drag. Moreover, although rebuilding efforts will start in Texas, we have found there to be considerable delays in prior hurricanes, suggesting 4Q may not benefit much from rebuilding. Looking ahead, reconstruction will create upside risk to growth in early 2018. The literature is mixed on the net impact on economic growth from natural disasters; it partly depends on the type of reconstruction and whether the changes improve upon the existing structure or facilities (boosts productivity) or simply restores them to their previous state.


We estimate the near-term effect to GDP growth using two approaches: (1) adjusting the assumptions for the high frequency data in our "nowcasting" GDP model and (2) back-of-the-envelope estimates based on the share of output from the hurricane-stricken areas. For the first, we create a counterfactual assuming no hurricane, which is based on prior monthly inputs to our GDP tracking estimates. We then plug in new forecasts for the remaining August and September data, relying on findings from our earlier analysis of the high frequency data. Using this approach, Harvey slices 0.4ppt from growth, owing to weaker consumer spending (mostly autos) and construction spending as well as a bigger inventory drawdown (Table 1). This is consistent with the range of estimates from our back-of-the-envelope estimate. Houston is the fourth most populous city in the US and the Houston metro area accounts for 3.2% of GDP (as of 2015). As such, a 5-15% decline in output over 3Q due to the hurricane would slice 0.2-0.5ppt from GDP growth.


Finally, and most unexpectedly, is how Harvey's impact streetched as far as Washington, DC, where as BofA states simply, "it helped"

Policymakers came back after the summer vacation with a number of mandates: extend the debt ceiling, fund the government past 30 September and provide funds for Hurricane Harvey relief. The desire to pass relief for Harvey has helped to generate a bipartisan plan to temporarily extend the debt limit and fund the government. On 7 September, the Senate passed a bill by a vote of 80-17 which provides $15.25bn in Hurricane relief (with provisions to address potential damage from Hurricane Irma), a continuing resolution and a suspension of the debt limit through 8 December. While this prevents a crisis in the coming weeks, it sets up for a battle in mid-December around the budget and the debt ceiling. Essentially Harvey pushed the potential fight around the budget and debt ceiling into the end of the year.