Value added by brokerages
Hi. I'm looking at the economic impact of brokerage services (sector 435) in a number of states. To pick an extreme example, let's look at Mississippi. I'm using the 2014 data.
For every $100 of revenue, sector 435 in Mississippi spends $120 on intermediate inputs from sector 435 itself (69% of that outside the state)! It spends $113 on noncomparable imports, $80 on real estate, $71 on other financial investment services, etc. The result is that, while there's $87 of employee compensation and $42 of proprietor income, other property income shows a $789 loss, and total value added is strongly negative.
The result of this is that impacts of industry changes in this sector in this state come out strongly negative in GDP terms.
Some of the same unusual characteristics are in the national data for this sector as well, albeit less starkly. OPI shows a 30% loss in the national data, although total value added for the industry is still positive, as it is in most states. In 2013 (R3), there was a 25% OPI loss at the national level.
Could you help me understand the strange economics of this industry in general (how is it so consistently losing money? Is there a strange accounting issue I'm not grasping?), and in states like Mississippi in particular?
Nationally, I assume the intermediate spend and value added data are grounded in BEA tables, but can you tell me where the Mississippi intermediate spend and value added numbers come from?
Thanks.
-
Hello, The large share of 435 making purchases from 435 is related to that sector’s spending pattern in the 5-year BEA input-output benchmark (BM). This data source provides the most industry detail available for industry spending patterns. Sector 435 includes NAICS codes 5231 and 5232, which includes investment banking, securities brokerages, commodity contracts brokerages, and security exchanges (see here: https://www.census.gov/cgi-bin/sssd/naics/naicsrch?code=523&search=2012%20NAICS%20Search). Given the mix of firm types, one would expect there to be significant money going to similarly-classified firms. Stock brokers typically pay transaction fees on exchanges, for example (indeed, this is one of the most important ways an exchange generates revenue). Brokerages may pay other brokerages fees, too, if using another brokerage’s services or an investment bank’s services to buy or sell assets. Investment banks also sometimes earn fees by operating their own exchanges. Note that transfers of assets, e.g., stocks, are not part of firms’ intermediate spending, but associated fees are part of intermediate spending. Every year, we obtain data on value-added components by state, at roughly the 3-digit NAICS level, from the BEA. This data series combines, conceptually, IMPLAN’s Other Property Income (OPI) category and Proprietor Income (PI) category into something called Gross Operating Surplus (GOS). The data are available at a 1-year lag relative to the IMPLAN data year (e.g., IMPLAN’s 2014 uses data for 2013 from the BEA, which IMPLAN projects). At the level of industry detail available from the BEA in this series, we know data only for NAICS 523, which maps to IMPLAN sectors 435 and 436. Based on the available data at the time, the BEA reported NAICS 523 in Mississippi to have GOS of $19 million in 2013 (the figure has since been revised to $13 million). Historically, GOS in this sector is estimated to be negative (financial losses) almost as often as it is positive. To divide GOS into its component OPI and PI parts, IMPLAN uses estimates of PI from the BEA in a similar series. BEA’s implied PI in 2013 in that sector in Mississippi, at the time of obtaining the raw data, was approximately $308 million. This implies a 2013 value for OPI of -$289 million, a substantially negative value. IMPLAN then needs to disaggregate and project BEA’s values into two different sectors, 435 and 436. IMPLAN uses Benchmark data to do this, along with current estimates of EC and PI. In the most recent Benchmark, the Benchmark-equivalent sector of 435 had negative GOS and 436 had positive GOS, but our estimated national control totals for OPI (a component of GOS, as noted above) in both sectors was negative. Since our OPI data by detailed sector (rather than the 3-digit level reported by BEA annually) is first estimated at the national level, with all sub-national places then being controlled to these national values, all places are going to have negative OPI in sector 435 unless the implied OPI for the state (derived from the 3-digit NAICS BEA data) is positive. Mississippi is particularly negative because of the significant negativity implied by the annual BEA 3-digit NAICS GOS data at the state level. Generally, the state-level intermediate expenditures (IE) are estimated after estimating all of the value added components and output. Output is estimated in a way so that IE/Output is targeted to be close to the national value, after adjusting for state-specific value added data. -
My apologies if I'm missing something obvious, but here's what I have for Mississippi for sectors 435 and 436: [table] [tr] [td][/td] [td]435[/td] [td]436[/td] [td]Total[/td] [/tr] [tr] [td]PI[/td] [td]$58M[/td] [td]$253M[/td] [td]$311M[/td] [/tr] [tr] [td]OPI[/td] [td]$-1,082M[/td] [td]$276M[/td] [td]$-806M[/td] [/tr] [tr] [td]GOS[/td] [td]$-1,024M[/td] [td]$529M[/td] [td]$-495M[/td] [/tr] [/table] Again, these are from the 2014 IMPLAN data, for a model constructed for Mississippi state in the online software. In other words, I'm not getting a GOS for these two sectors of $19 or $15 million, but of $-495 million. For PI, I get $311 million, which is very close to the $308 million you quote. Can you tell me what's going on? Thanks! -
Hello, We apologize for the delay. The calculations for GOS, OPI, and PI at the state level are meant primarily as an illustration that the values can be substantially negative. We perform the same calculations for each state, but then need to control those state values to the national control total for OPI. The national OPI value is estimated, in part, to try to preserve the relationship between IE/Output ratio for the year and the benchmark value. For our 2014 data, this entailed multiplying negative state OPI values in sector 435 by approximately a factor of 4. If our estimate of national OPI for this sector had been any less negative than it is (approximately -$44 million), the national IE/Output ratio would have strayed farther from the benchmark data on IE/Output.
Please sign in to leave a comment.
Comments
4 comments