Sale Indices
IVG-R
The IVG-R (Índice de Valores de Garantia de Imóveis Residenciais) is published by the Brazilian Central Bank. It covers the national market from 2001 and is widely regarded as the official RPPI in Brazil. The index is based on median appraisal prices from mortgage contracts, smoothed with a 3-month moving average and the HP filter.
ivgr <- get_dataset("rppi", "ivgr")
glimpse(ivgr)
#> Rows: 301
#> Columns: 5
#> $ date <date> 2001-03-01, 2001-04-01, 2001-05-01, 2001-06-01, 2001-07-01, …
#> $ name_geo <chr> "Brazil", "Brazil", "Brazil", "Brazil", "Brazil", "Brazil", "…
#> $ index <dbl> 100.00, 100.08, 100.15, 100.22, 100.29, 100.36, 100.43, 100.4…
#> $ chg <dbl> NA, 0.0008000000, 0.0006994404, 0.0006989516, 0.0006984634, 0…
#> $ acum12m <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 0.00980000, 0…ggplot(ivgr, aes(date, index)) +
geom_line(color = color_palette[1], linewidth = 0.7) +
labs(
title = "IVG-R — National Sale Index",
x = NULL,
y = "Index"
) +
theme_series()
IGMI-R
The IGMI-R (Índice Geral do Mercado Imobiliário Residencial) is published by Abecip/FGV. It uses a hedonic pricing model applied to actual transaction prices, making it the most methodologically rigorous city-level index available. Coverage spans major Brazilian cities from 2014.
igmi <- get_dataset("rppi", "igmi")
glimpse(igmi)
#> Rows: 1,617
#> Columns: 5
#> $ date <date> 2014-01-01, 2014-01-01, 2014-01-01, 2014-01-01, 2014-01-01,…
#> $ name_muni <chr> "São Paulo", "Rio De Janeiro", "Belo Horizonte", "Fortaleza"…
#> $ index <dbl> 100.00000, 100.00000, 100.00000, 100.00000, 100.00000, 100.0…
#> $ chg <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 0.0003654000, 0.…
#> $ acum12m <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …main_cities <- c("São Paulo", "Rio De Janeiro", "Belo Horizonte", "Brasília")
subigmi <- igmi |>
filter(name_muni %in% main_cities)
ggplot(subigmi, aes(date, index, color = name_muni)) +
geom_line(linewidth = 0.8) +
labs(
title = "IGMI-R — Sale Index by City",
x = NULL,
y = "Index",
color = NULL
) +
theme_series()
FipeZap (sale)
FipeZap is the broadest RPPI in terms of geographic coverage, with indices for more than 50 cities. It is based on median listing prices rather than transaction prices. Available from 2008 for both residential and commercial markets, with breakdowns by number of bedrooms. Prices are stratified by region and income level, using Census weights, and the final series is smoothed using a 3-month moving average.
Key columns specific to FipeZap:
market(residential/commercial),rent_sale(rent/sale),rooms(1, 2, 3, 4+, total),variable(index, chg, acum12m, price_m2, yield).
fz <- get_dataset("rppi", table = "fipezap")
glimpse(fz)
#> Rows: 677,160
#> Columns: 7
#> $ date <date> 2008-01-01, 2008-01-01, 2008-01-01, 2008-01-01, 2008-01-01,…
#> $ name_muni <chr> "Brazil", "Brazil", "Brazil", "Brazil", "Brazil", "Brazil", …
#> $ market <chr> "residential", "residential", "residential", "residential", …
#> $ rent_sale <chr> "sale", "sale", "sale", "sale", "sale", "sale", "sale", "sal…
#> $ variable <chr> "index", "index", "index", "index", "index", "chg", "chg", "…
#> $ rooms <chr> "total", "1", "2", "3", "4", "total", "1", "2", "3", "4", "t…
#> $ value <dbl> 41.81107, 39.46609, 40.45048, 43.47948, 47.09432, NA, NA, NA…Working with this dataset requires more complex filtering. For most
use cases, however, the main choices are selecting the
market (either ‘residential’ or ‘commercial’),
rent_sale (either ‘sale’ or ‘rent’), and the desired
variable.
subzap <- fz |>
filter(
market == "residential",
rent_sale == "sale",
rooms == "total",
variable == "index",
name_muni %in% main_cities
)ggplot(subzap, aes(date, value, color = name_muni)) +
geom_line(linewidth = 0.8) +
labs(
title = "FipeZap — Residential Sale Index",
x = NULL,
y = "Index",
color = NULL
) +
theme_series()
Comparing indices
As mentioned above, the table parameter accepts a
"sale" argument, which stacks all available indices into a
single tibble.
sale_indices <- get_dataset("rppi", "sale")
glimpse(sale_indices)
#> Rows: 14,458
#> Columns: 6
#> $ source <chr> "IGMI-R", "IGMI-R", "IGMI-R", "IGMI-R", "IGMI-R", "IGMI-R", …
#> $ date <date> 2014-01-01, 2014-01-01, 2014-01-01, 2014-01-01, 2014-01-01,…
#> $ name_muni <chr> "São Paulo", "Rio De Janeiro", "Belo Horizonte", "Fortaleza"…
#> $ index <dbl> 100.00000, 100.00000, 100.00000, 100.00000, 100.00000, 100.0…
#> $ chg <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 0.0003654000, 0.…
#> $ acum12m <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …For convenience, the stacked version of the dataset implicitly only
includes rooms == "total" for FipeZap. Also, “Brazil”
(spelled with a ‘z’) is listed under name_muni, which shows
the nation-wide index, instead of the city specific ones.
The plot shows the three sale indexes at a national level. While there’s some convergence between the indices, there is a growing gap between them in the post-pandemic years. This highlights the importance of choosing the right index for your use case.
ggplot(comp_index, aes(date, acum12m * 100, color = source)) +
geom_line(linewidth = 0.7) +
labs(
title = "Comparing Sale Indices in Brazil",
subtitle = "12-month accumulated change (%)",
y = "%",
color = NULL
) +
theme_series()
