************************** “There are three kinds of lies: lies, damned lies and statistics.” Benjamin Disraeli
Statistics without informed context are usually worthless, easily manipulated and often misleading. One can make virtually any case -- positive or negative -- by choosing a single average or median statistic relating to a short period of time and a small data set, and then cherry picking what you’re comparing today’s data to (last month, last year, or the peak of the market). Conversely, too large a data set may be misleading: the overall national trend may misrepresent California’s, and the state’s can be different from the Bay Area’s, the Bay Area’s from the city’s, and within San Francisco itself, distinct neighborhoods are often different markets going in significantly different directions. In particular, absent some huge economic event, such as the September 2008 financial markets meltdown, monthly fluctuations in median home sales prices are usually meaningless. Median prices often fluctuate up and down within a 5 to 10% range from one month to the next, even in stable markets.
One can only be sure market values are trending up or down if that trend is consistent over the longer term, minimally 4 to 6 months. Any definitive trend in prices and values should also be reflected in other market statistics such as average dollar per square foot, days on market, months’ supply of inventory, percentage of listings accepting offers, percentage of distress sales, and so on.
When assessing market changes calculated by computerized algorithms using very general data sets – such as Case Shiller’s or Zillow’s -- one should be clear on the details. For example, the Case Shiller Index for “San Francisco” reflects an analysis of a “metro area” comprising 5 counties with wildly varying markets (Pinole to Pacific Heights). And for the city of San Francisco, one should look at the Case-Shiller “High Tier” price Index, not the general Index. It also makes sense to assume a sensible margin in error. As an egregious example, Zillow’s property valuations usually build in a 10-25% margin of error on either side of their “Zestimate” of value. A 1-3% value change indicated by the Case Shiller overall home Index for the SF metro area, then applied by a commentator to condo values in SOMA or house values in the Marina, should be taken with a grain of salt.
Always look for consistent, longer term trends across a wide range of market quantifying statistics.
MEDIAN SALES PRICE is that price at which half the sales occur for more and half for less. It can be, and often is, affected by other factors besides changes in market values, such as short-term or seasonal changes in inventory or buying trends. The median sales price for homes (in all their infinite variety) is not like the price for a share of stock (all the same), and monthly fluctuations in median price are generally meaningless. If market values are truly changing, the median price will consistently rise or sink over a longer term than just 2 or 3 months, and also be supported by other supply and demand statistical trends.
AVERAGE SALES PRICE is calculated by adding up all the sales prices and dividing by the number of sales. It is different from median sales price, but like medians, averages can be affected by other factors besides changes in value, such as fluctuations in average unit size. Averages may also be distorted by a few sales that are abnormally high or low, especially when the number of sales is low. Average sales prices are usually higher than median sales prices.
DAYS ON MARKET (DOM) are the number of days between a listing going on market and accepting an offer. The lower the average days on market figure, typically the stronger the buyer demand and the hotter the market. Note that this statistic is distorted by distress sales, which often have a very high DOM, by that minority percentage of listings that sell after multiple price reductions, and by deals that fall through after offer acceptance (the listings come back on market, but the DOM clock keeping ticking). Appealing, well-priced new listings often accept offers within 7 to 14 days of coming on market.
MONTHS SUPPLY OF INVENTORY (MSI) reflects the number of months it would take to sell the existing inventory of homes for sale at current market conditions. The lower the MSI, the stronger the demand as compared to the supply and the hotter the market. Typically, below 3-4 months of inventory is considered a "Seller's market", 4-6 months a relatively balanced market, and above 6 months, a "Buyer's market."
DOLLAR PER SQUARE FOOT ($/sqft) is based upon the home’s interior living space and does not include garages, unfinished attics and basements, rooms built without permit, lot size, or patios and decks -- though all these can still add value to a home. These figures are usually derived from appraisals or tax records, but are sometimes unreliable or unreported altogether. Generally speaking, about 60-80% of listings report square footage and dollar per square foot averages are calculated on these listings alone. All things being equal, a house will sell for a higher dollar per square foot than a condo (due to land value), a condo higher than a TIC (quality of title), and a TIC higher than a multi-unit building (quality of use). Everything being equal, a smaller home will sell for a higher $/sqft than a larger one. (However, things are rarely equal in real estate.) There are often surprisingly wide variations of value within neighborhoods and averages may be distorted by one or two sales substantially higher or lower than the norm, especially when the total number of sales is small. Location, condition, amenities, parking, views, lot size & outdoor space all affect $/sqft home values. Typically, the highest dollar per square foot figures in San Francisco are achieved by penthouse condos with utterly spectacular views in prestige buildings.
SAN FRANCISCO REALTOR DISTRICTS
District 1: Sea Cliff, Lake Street, Richmond (Inner, Central, Outer), Jordan Park/Laurel Heights, Lone Mountain
District 2: Sunset & Parkside (Inner, Central, Outer), Golden Gate Heights
District 3: Lake Shore, Lakeside, Merced Manor, Merced Heights, Ingleside, Ingleside Heights, Oceanview
District 4: St. Francis Wood, Forest Hill, West Portal, Forest Knolls, Diamond Heights, Midtown Terrace, Miraloma Park, Sunnyside, Balboa Terrace, Ingleside Terrace, Mt. Davidson Manor, Sherwood Forest, Monterey Heights, Westwood Highlands
District 5: Noe Valley, Eureka Valley (Castro, Liberty Hill), Cole Valley, Glen Park, Corona Heights, Clarendon Heights, Ashbury Heights, Buena Vista Park, Haight Ashbury, Duboce Triangle, Twin Peaks, Mission Dolores, Parnassus Heights
District 6: Hayes Valley, North of Panhandle (NOPA), Alamo Square, Western Addition, Anza Vista, Lower Pacific Heights
District 7: Pacific Heights, Presidio Heights, Cow Hollow, Marina
District 8: Russian Hill, Nob Hill, Telegraph Hill, North Beach, Financial District, North Waterfront, Downtown, Van Ness/ Civic Center, Tenderloin
District 9: SoMa, South Beach, Mission Bay, Potrero Hill, Dogpatch, Bernal Heights, Inner Mission, Yerba Buena
District 10: Bayview, Bayview Heights, Excelsior, Portola, Visitacion Valley, Silver Terrace, Mission Terrace, Crocker Amazon, Outer Mission
Some Realtor districts contain neighborhoods that are relatively homogeneous in general home values, such as districts 5 and 7, and others contain neighborhoods of wildly different values, such as district 8 which includes both Russian Hill and the Tenderloin.
