The NCREIF Property Index is a quarterly time series composite total rate of return measure of investment performance of a very large pool of individual commercial real estate properties acquired in the private market for investment purposes only. All properties in the NPI have been acquired, at least in part, on behalf of tax-exempt institutional investors - the great majority being pension funds. As such, all properties are held in a fiduciary environment.
Properties in the NPI are accounted for using market value accounting standards. Data contributed to NCREIF is expected to comply with REIS. Because the NCREIF Property Index measures performance at the property level without considering investment or capital structure arrangements, information reported to the index will be different from information reported to investors. For example, interest expense reported to investors would not be included in the NCREIF Property Index. However, because the property information reported to the index is expected to be derived from the same underlying books and records, because it is expected to form the underlying basis for investor reporting, and because accounting methods are required to be consistent, fundamentally consistent information expectations exist.
NCREIF requires that properties included in the NPI be valued at least quarterly, either internally or externally, using standard commercial real estate appraisal methodology. Each property must be independently appraised a minimum of once every three years. Because the NPI is a measure of private market real estate performance, the capital value component of return is predominately the product of property appraisals. As such, the NPI is often referred to as an "appraisal based index."
The qualifications for inclusion in the NPI are:
The composition of the NPI can change over time. The numbers of properties changes as Data Contributing Members buy and sell properties and new Data Contributing Members are added. Properties exit the NPI when assets are sold or otherwise leave the database. All historical data remains in the database and in the Index. The Index represents investment returns from a single class of investor. As such, the NPI may not be representative of the market as a whole.
There are three requirements that must be met for initial entrance of a property into the NPI. The requirements are as follows: 1) The property must be classified as one of the following property types: office, retail, industrial, apartment, or hotel. 2) The accounting of the property must be performed using market value accounting (see "What Are The Accounting Policies for the NPI") 3) The property must be an operating property
a) For a newly developed property, operating is defined as reaching 60% occupancy or having been available for occupancy for a year from its certificate of occupancy (CO). b) If a property has been recently purchased with a "redevelopment" strategy, and the property is undergoing substantial expansion or re-tenanting, rehabilitation or remodeling, the property is defined as operating when occupancy reaches 60%. c) All existing properties (not recently developed or undergoing redevelopment as covered in a) or b) above) that are purchased regardless of current occupancy are defined as operating properties.
Once a property has met all the criteria to be included in the NPI it is only removed if:
If a property is sold from one NCREIF member to another NCREIF member, the property leaves the NPI as a sale, but then re-enters the NPI a quarter later (the first full quarter of ownership) if, at that time, the property meets all of the inclusion criteria for the NPI.
If there is a change in use for the property included in the NPI, e.g., conversion to a different property type, the property will normally be classified under its original property type until the property is deemed substantially converted to the new property type. If the conversion is to a property type that still qualifies for inclusion in the NPI then it will stay in the NPI but move to the new property type upon substantial completion. If the property is converted to a use that does not qualify for inclusion in the NPI then it will be removed from the NPI upon substantial completion of the use change, although its history will remain in the index.
If a property is being converted into condominiums by the manager, the proceeds from sale of the condominium units are reported as partial sales. The property should continue to be classified under its original use until the units have been sold.
NCREIF has no official position regarding the classification of the NPI as being a "core" investment strategy index.
The NPI is a collection of unlevered property level returns that meet NCREIF's defined inclusion requirements. Some consider it reflective of a "core" investment strategy based on the property types included, the exclusion of development projects, and returns being reported on an unleveraged basis. Others, meanwhile, note the need to consider geography, leverage, and credit risk, to name a few, when defining a "core" investment strategy. The NPI includes properties that meet its defined inclusion criteria regardless of the location, risk of existing leases, and potentially varying occupancy levels.
The Index is set at 100 starting fourth quarter of 1977. Calculations are based on quarterly returns of individual properties before the deduction of portfolio-level management fees, but inclusive of property level management fees. Each property's return is weighted by its market value (value-weighted). Index values are calculated for income, capital value and total.
The NPI quarterly, annual and annualized total returns consist of three components of return - income, capital and total.
The Income Return measures that portion of total return attributable to each NPI property's net operating income, or NOI. Net operating income (NOI) is gross rental income plus any other income less operating expenses - utilities, maintenance, taxes, property management, insurance, etc. The income return is computed by dividing NOI by the average daily investment for each quarter. The formula takes into consideration any capital improvements and/or any partial sales that occurred during the quarter.
The Capital Return measures the change in market value from one period to the next. A property's value can go up (appreciation) or it can decline (depreciation) depending on market forces. The formula takes into consideration any capital improvements and/or any partial sales that occurred during the quarter. When a property enters the Index, the capital return is not impacted until the second quarter of inclusion.
The NPI can be thought of as a market cap weighted index as opposed to an equal weighted index. Thus, larger properties in terms of market value have a greater impact on the index than smaller properties. What is being calculated is essentially the return for the entire portfolio of NCREIF properties.
In equal weighted performance analysis each property counts the same weight, regardless of the value of the fund. Index aggregates are typically value-weighted. Equal weighting however; is useful for statistical analysis as well as peer comparisons. Some of the NPI query tools and Fund Level Indices provide Equal-Weighted Return Analysis, however the overall index is published on a Value-Weighted Basis.
An NPI quarter-to-quarter return provides an estimate of the quarterly Internal Rate of Return (IRR) as if a property was purchased at the beginning of the quarter and sold at the end of the quarter with the investor receiving all net cash flow (NOI-Cap X) during the quarter.
The NPI rate of return formula assumes:
These assumptions are the reason that the formulas above include terms in the denominator for adding ½ Capital Expenditures, subtracting ½ Partial Sales and subtracting 1/3 NOI. These adjustments make the denominator an estimate of the average investment during the quarter.
This methodology is well accepted in performance measurement and is consistent with what is referred to as the "Modified Dietz" formula for measuring investment performance which considers the average daily investment in an asset. In this case it is as if the NOI was received on days 30,60 and 90 of the quarter and capital expenditures and partial sales occur on day 45.
The index discussed earlier is calculated from the quarterly returns by increasing the index level by the return for each quarter. The NCREIF Property Index is started with ha value of 100 in the fourth quarter of 1977. Each quarter the index represents the cumulative return of the periods following. The index explains what the value of $100 invested in the NCREIF property set from 1977 would be in the index period. This can be thought as chain linking the quarterly returns.
An index of total return, capital return and income return are calculated from NCREIF. The cumulative time-weighted return between any two periods can be calculated easily from this index by a simple change in index value calculation (Index Value Period 2 - Index Value Period 1) divided by Index Value Period 1.
Two sets of data are collected for the NPI. The first represents property specific descriptor information submitted when a property enters the database for the first time. Key property data fields include: Property Name, Address, Zip Code Property Type & Sub-Type Acquisition Date Investment Cost Number of Floors, Units (Apartments, Hotels) Gross and Net Rentable Square Feet Year Built Date of Most Recent Renovation Occupancy Rate Retail Anchor Store Square Feet Retail Non-Anchor Store Square Feet How Acquired Joint Venture - Note, on JV properties, the data is reported to NCREIF as if 100% owned.
The second data set is collected quarterly and includes the components of return needed to calculate quarterly rates of return and index values. NCREIF collects considerably more data than what is required to calculate the NPI. Additional data are used in NPI data validation tests, to calculate additional statistical measures of performance, to develop operating benchmarks and for use in real estate research. The key quarterly data fields include: Market Value (MV) Gross Revenues Breakouts Operating Expenses Breakouts Net Operating Income (NOI) Capital Expenditures (Cap. X.) Partial Sales Proceeds Appraisal Activity Percentage Leased Percentage Occupied Leaseable Area Leasing Commissions Tenant Improvements
Data are submitted in accordance with NCREIF's data submission manual, NCREIF Property Indexes Data Collection and Reporting Procedures. The manual is a detailed step-by-step narrative, which clearly identifies all data components required for submission and instructions on how to deliver data to NCREIF. The manual is periodically updated to reflect any additional data submission requirements and/or whenever technological improvements in the information delivery systems are made.
The data NCREIF collects originates from the accounting and property management systems of Data Contributors. The process involves uploading data files to a NCREIF standard data submission template through a secured members-only area at NCREIF's web site. Data are then extracted from each Data Contributor's files and downloaded into NCREIF's database server. The server is under the operation and control of NCREIF's Database Manager.
Data are maintained in a state of confidentiality. Individual property data are masked in order to protect each Data Contributor's proprietary property level information. Individual property data cannot be accessed without Data Contributor approval.
Once the quarterly data submission process is completed and the NPI and other standard quantitative products have been published, the data are transferred to a second secured database for web site queries. All data are available for research. To protect the identity of individual property level data, the minimum number of properties accessible for research is four and must be owned/managed by two or more Data Contributors.
NCREIF Data Contributing Members are required to deliver all NPI qualifying data no later than 20 days following the end of the most recent business quarter. Note: NCREIF operates on a January-December calendar-year. If a member is late in contributing data to NCREIF and the result is a delay in the release of the Index, the member's name is published in the Index report. This has only happened one time since this rule was established in 2000. In the case of a late member submission, NCREIF staff will work with the manager to try to help resolve any issue(s) that might cause a delay and if a delay is inevitable, the Index Oversight Committee (IOC) will have to make a decision to wait a day or two in hopes of receiving the data or to publish without the member's data. The size of the manager will have a large impact on this decision.
NCREIF Staff meets with the Index Oversight Committee (IOC) via conference call 25 days following the end of the business quarter to review the data and approve the release of the NPI.
NCREIF produces various data products once the index is released, a description of those products are attached. Most products are released the day the index is approved and released.
NCREIF has set up a three-step data validation process. It is designed to identify and correct data submission errors. Data are delivered to NCREIF via a standard data submission software package specifically designed for this process. Built into the system of the Data Contributors are validation tests run on a pre-submission basis to indicate any data errors or other warnings.
Errors are defined as problems needing fixing before final submission. Warnings are defined as outliers evidencing deviations from standard norms. Warnings require being checked into before final submission. Typically, a warning is issued when an unusually large change in an individual property's market value takes place from one quarter to the next. Most often, a big change in value is the result of a property sale, the net proceeds of which deviate significantly from its most recent appraised value.
Data Contributors represent the first level of validation screen. A significant majority of data errors and outliers are flagged at the front end of the submission process. Individual Data Contributor systems include NCREIF compatible software to run validation tests. Data Contributors are in the best position to identify and correct data errors.
When the data are delivered to NCREIF, NCREIF conducts its own level of validation. This is a two-step process. The first involves running the same validation tests as run by the Data Contributor. The second involves comparing a Data Contributor's returns with those of all Data Contributors. This test produces a summary report of potential errors and outliers, which is then sent to each Data Contributor flagged for review. When a Data Contributor is notified of possible data errors or warnings, the Data Contributor reviews the report, makes whatever corrections are necessary, resubmits the corrected data, and then signs off on the accuracy of the returns reported for the quarter.
Examples of validation rules that are performed include examining properties with the greatest increase or decrease in value, properties with the highest or lowest income, capital or total returns, properties with a price per square foot that is outside of the normal range for peers of the same property type, excessive capital expenditures, partial sales that are a high percentage of the property value, and other similar tests for any input that differs significantly from the norm.
A third level of data validation is conducted by NCREIF's Index Oversight Committee (IOC). At the end of each data reporting period and prior to the release of the NPI quarterly returns, the IOC reviews property type and regional returns for purposes of comparing current NPI returns with industry forecasted returns based on current market conditions. Any data concerns raised by the IOC are immediately checked and, if necessary, confirmed or corrected via Data Contributor follow up.
IOC responsibilities include reviewing and recommending changes to the Board of Directors in index policies and procedures, required data elements, qualifying properties, quarterly index production review and the taking of any other actions deemed necessary to assure index statistical integrity.
The IOC is also responsible for reviewing and recommending to the Board for approval any changes to the Index Policies and Procedures, including, but not limited, to sanctions for failure to report data in a timely, thorough and accurate manner, all required data elements and qualifying properties.
The IOC is made up of 5 senior researchers from member firms that have a significant understanding of the database, the index and its methodologies, and also an understanding of the current markets. A majority vote is technically required to approve and release the index or to make recommendations to the Board of Directors on policy changes or other such items.
The NPI was "frozen" each quarter beginning First Quarter 2003. This means that a snapshot of the index was taken each quarter and changes are not made historically unless there is a significant error that is caught later that would require the restatement of the NPI. A significant error is considered to be one that affects the National, Property Type or Regional returns by 10 basis points or greater. A restatement has only happened once since the "freeze." The restatement took place in Fourth Quarter 2004 due to a manager submission error that affected a property type return by 16 basis points.
Prior to 2003, the NPI was restated historically every quarter for various reasons, the majority being the fact that when there were property sales in a given quarter, the previous quarter's ending market value was changed to reflect the sales price.
Changes in data were also due to the fact that when a new member joined NCREIF, their historical data was put into the database and would be included in the NPI returns on future releases. There were also data errors that were caught after the release of the index and those would be corrected historically, though over time errors became less and less frequent as new and better validation screens were built into the systems and as information technology programs were enhanced and upgraded to catch potential errors during the submission process.
Beginning First Quarter 2003 NCREIF made a decision to "freeze" the NPI for two reasons. First, the database had reached a size, in both number of properties and market values, that the benefits of backdating were no longer important. Secondly, more and more investment managers and in-house real estate staffs' compensation is being tied to outperforming the NPI. The market wants a single return number at the time when NCREIF releases the quarterly NPI.
Each quarter, the Data Contributing Members submit a market value for each NPI qualifying property. The value the Data Contributor submits is the value they believe is the property's fair market value as of that particular reporting period - i.e., the end of each calendar quarter.
A change in value from one quarter to another can be for one of several reasons. The property was externally appraised by an independent third party appraiser. Observed changes in market conditions as so determined by the manager to recognize any changes during the quarter in occupancy, rental rates, capitalization rates, interest rates, a partial sale, unexpected capital expenditures, or changes in discount rates. A property value may be adjusted only for capital expenditures made during the quarter - effectively, an accounting adjustment to reflect the amount of the capital expenditure. The value submitted can be the previous quarter's value because, in the judgment of the manager/owner, the property's value did not change during the period.
An appraisal is an estimate of a property's current market value at a particular point in time based on traditional appraisal methodology and techniques. Appraisers take into account recent sales of comparable properties, replacement cost, and other appraisals of similar type assets, when such information is available.
Whenever an NPI property is appraised, Data Contributing Members are required to report as to whether the appraisal was internal or external. An internal appraisal is one prepared at the investment manager or institutional investor level by the in-house professional real estate staff. An external appraisal is one performed by an independent party, typically an MAI (Member of the Appraisal Institute).
The NPI database consists of all qualifying properties whose returns represent the most recent quarterly NPI calculation, plus all properties that have been in the database historically, but are no longer included in the Index.
Of the 12,455 properties that have been in the Index, 5,116 with a combined net sales proceeds generated of $104.9 billion represent outright sales. When a property is sold, the Data Contributor submits to NCREIF the net sales proceeds received, which represents the final quarterly value for that particular property. When a sale takes place, the property "exits" the Index. While the sold property no longer remains in the NPI, its history is a permanent part of the NPI's historical returns. More properties leave the NPI database as the result of outright sales than for any other reason.
An "Event" property exits the database for reasons other than an outright sale. These include when: a pool of properties is sold as a portfolio; a Data Contributor is merged or consolidated with another manager; a Data Contributor is terminated or goes out of business; a property was destroyed by an act of God; a property was transferred to a lender or joint venture partner.
Two "Events" cause properties to exit the database more than all others combined. The first is a portfolio sale. A portfolio sale occurs where a manager sells a number of properties to a single buyer in a single transaction. When this happens, while technically a sale, portfolio sales prices are not allocated on a property-by-property basis. Under these circumstances, the individual properties that make up the portfolio exit the NPI database at their respective most recent market values.
"Event" properties also represent properties transferred from one manager to another as a result of a merger, consolidation or termination. When such an event takes place, the property exits the database at its most recent appraised value. If the property is then transferred to another Data Contributing Member, the property re-enters the database under the name of the successor manager.
The NPI is not, and was never intended to be, a static database. It is both dynamic and evolving. Each quarter, the database property mix changes due to:
New property purchases Property sales Changes in NCREIF data Contributing membership "Events," such as portfolio sales, manager mergers, terminations
Annual and Annualized Returns are computed by chain linking quarterly rates of return to calculate time-weighted rates of return for annual and annualized periods under study. Returns are assumed to be compounded quarterly. For periods of over one year, returns are expressed on a return-per-year basis.
Compounded quarterly. Quarterly returns are compounded in annual calculations. This assumes that returns in early quarters are "reinvested" in the property in subsequent quarters. An annual"compounded" return is calculated from quarterly returns as follows: Example Quarter 1 return = 2.3607% Quarter 2 return = 2.4737% Quarter 3 return =1.5957% Quarter 4 return = 2.4734% Annual Return is equal to ((1+Quarter 1 return) x (1+Quarter 2 return) x (1+Quarter 3 return) x (1+Quarter 4 return)) - 1 The annual return would therefore be (1.023607 x 1.024737 x 1.015957 x 1.024734) - 1 = 9.2024%.
For multi-year periods, or those shorter than a year, the above formula would show an annualized rate of return for the calculation period.>
Click here for a primer on NCREIF Derivatives
Members are required to submit data from all assets under management. Often this data includes Non-NPI Qualifying properties. These properties can be non-Qualifying for the NPI for a number of reasons including the tax status of the ownership, appraisal policy, property type other than those in the NPI or life cycle of the property (i.e. under development or in lease-up). Additional members have joined to contribute non-NPI qualifying property including REITs, Insurance Companies and other taxable institutional real estate investors. The combined npi-qualifying and non-qualifying data is available for research purposes through the operating data product.
The operating data includes a broader set of properties than those included in the NPI. Some property data is not eligible for the NPI due to the valuation policy, ownership structure or the like, however is included within the operating data. This allows for a more robust sample of line item income and expense data.
The NCREIF operating data includes the income and expense categories. This allows you to see trends in different income and expense categories for selected property types and locations. Each user has their own private query list in which he or she can store queries for later review, edit queries, or remove old queries.
The detailed income and expense data began collection in 2000.
Base Rent - Income derived strictly from leased space, i.e., not including Percentage Rents (contingent rents), which is a separate data field.
Contingent Rent or Percentage Rent - Income received as a percentage of the gross or net income to a tenant. Most often applied in retail property leases.
Income from Reimbursements - Any payment made by tenants to reimburse the landlord. Generally these are reimbursements for expenses that have occurred which the landlord initially paid and is being reimbursed for.
Other Income - All other income going into the Net Operating Income calculation that is not otherwise categorized.
Total Income or Operating Revenue - Total Income is the sum of base rent, Contingent Rent, Income from Reimbursements and other income
Utility Expense - All utility costs, including water, sewer, power, fuel oil, etc.
Property Tax Expense - Property Tax includes Real Estate Taxes as well as personal property taxes on the investment.
Insurance Expense - Premiums paid to insure the property.
Repairs and Maintenance - All expenses associated with the daily upkeep of the property including materials and labor, maintenance salaries, cleaning and janitorial, engineering costs. This category does not include major capital expenses including building improvements, building expansions and other capital improvements (see description).
Marketing/Advertising - Marketing costs, including salaries, advertising, and promotions.
Property Management Fees - Fees paid to an external property manager, along with uncapitalized leasing agent fees.
General and Administrative Expenses - Office expenses, licenses, legal, travel, audit. Excludes amounts allocated to Repairs and Maintenance, Marketing/Advertising, or Property Management Fees, which are separate expense fields, as noted above.
Other Expenses - All other expenses accrued during the quarter going into the Net Operating Income calculation that is not otherwise categorized.
NOI (Net Operating Income) - NOI or Net Operating Income is the sum of the income on the property less the sum of the expenses (excluding capital expenses).
Tenant Improvement - Tenant Improvements include costs to construct (or reconstruct) space for tenant occupancy after the Certificate of Occupancy has been issued. Demolition costs (even when a tenant is not under lease), are also included in this category.
Leasing Commissions - Leasing commissions are fees paid for the procurement of tenants after the Certificate of Occupancy Date for the property?
NCREIF currently publishes two “cap rate" (or capitalization rate) series in the Property Value Trends (PVT). Both of these cap rates series are calculated by dividing the Net Operating Income by Market Value. The series only differ in the sample of properties included. Because accounting data is used, the most recent full quarter's income from the property, annualized, provides the net operating income. This would be different from most reported "cap rate" reports, which often use a forecast of year one net operating income. The "transaction cap rate" series is the Net Operating Income (defined above) divided by the transaction price of only properties that have sold from the NCREIF Property Index. The "current value cap rate" series is similar, but includes all properties that have been revalued in the quarter (so the market value is known to be current). The Current Value Cap Rate series is available at the regional level and property type level.
The Leveraged Property quarterly, annual and annualized total returns consist of three components of return - income, capital appreciation and total.
The Leveraged Property Income Return measures that portion of total return attributable to each NPI property's net operating income after interest expenses. The income return is computed by dividing NOI by the average daily investment for each quarter. The formula takes into consideration any capital improvements and/or any partial sales that occurred during the quarter.
Leverage Property Capital Appreciation Return: Measures the change in market value adjusted for any capital improvements, partial sales, principal repayment and other debt receipts or payments for the quarter.
Leverage Property Total Return: Includes appreciation (or depreciation), realized capital gain (or loss) and income. It is computed by adding the Income return and Capital Appreciation return on a quarterly basis.
The Index is set at 100 starting fourth quarter of 1982. Calculations are based on quarterly returns of individual properties before the deduction of portfolio-level management fees, but inclusive of property level management fees. Each property's return is weighted by its market value (value-weighted). Leveraged property index values are calculated for income, capital value and total.
It is an index based on properties that were in the NCREIF Property Index (NPI) and were sold that quarter. The index does not replace the NPI! It is a complimentary index to the appraisal based NPI. A transaction based index is often considered to be more comparable to stock and bond indices that are transaction based.
The NCREIF TBI (NTBI) is calculated by taking the average ratio of current sale price divided by a two quarter lagged appraisal (from the NPI database) among all the sold properties each quarter. That ratio is multiplied times the NPI cumulative capital appreciation index level, to convert the result into a transaction price index. The two quarter lagged appraisal is used instead of the current quarter’s appraisal in order to obtain an NPI reported valuation that is independent of transaction price. Studies have shown that an appraisal done two quarters prior to the sale of a property is not influenced by the subsequent transaction. Any appraisal done closer to the actual transaction may influence the appraisal price.
The diagram below shows a visual representation of how the NTBI is calculated.
The MIT TBI was calculated using a hedonic regression model.
The results from the two calculations are very similar, but the NCREIF version is simpler, more transparent and easier to understand. By not using a regression model, the results will be easier for users to explain and NCREIF can create additional indices with the data.
The NPI is a value weighted index composed of a capital appreciation return and an income return. It is calculated using appraised values. The TBI is an equal-weighted transaction and appraisal index. Neither one is right or wrong. They are two different methods of creating indices to track the real estate market. A value weighted index treats the NPI as a population or portfolio of all properties held by members whereas an equal weighted index views the properties in the NPI as a sample from a broader population of commercial real estate.
The NPI includes hotels. There is no TBI for the hotel property type, given the smaller population of properties and transactions in the NCREIF database. This may change in the future, but the current index series only includes apartment, industrial, office and retail property type indexes.
We use the equal-weighted version in the NTBI because we view the NTBI as a statistical estimation, based on a sample of properties (the ones that sold) inferring from that sample the change in value in the market represented by the entire NCREIF population of properties. We view each sold property as an equally-valid representative of that population. Hence, it makes sense for us to use the equal-weighted version of the NPI for comparison purposes. Also, the equal-weighted NPI tends to have less volatility (in effect it is more diversified, less dominated by a few large properties). Over time there should be no material difference in the trends, but there may be quarterly fluctuations that cause divergence depending on which properties transact.
The NTBI is not intended to be a benchmark. Only a small subset of properties in the NPI sells each quarter and thus do not represent the larger commercial real estate universe. Therefore, using the NTBI as a benchmarking index would not be appropriate. Furthermore, managers typically calculate the returns on their own funds using appraised values so they should compare these returns with an appraisal based index.
No. The national, all-property type ratio is multiplied times the individual property type or regional index. This is based on the assumption that the “appraisal lag” is systematic and does not vary significantly by property type or region. This assumption allows us to use all the properties to calculate the ratio and eliminate the noise that would result from basing the ratio on a single property type or region.
These indices still require an econometric model to estimate and are much more difficult to interpret. Thus, NCREIF has decided to not produce these indices.
However, NCREIF may make the data available to the academic community for computation of these indices in the future.
Yes, a new historical series will be created to reflect the change in methodology.
The negative autocorrelation in the NTBI is caused by "noise" in the transaction prices. Any given transaction is unique between essentially one unique buyer and one unique seller regarding one unique property. As a result, there is a certain amount of random dispersion around some theoretical "true" value (that no one can observe). From a statistical perspective it is like the dispersion in a random sample. If there are 100 balls in a jar each with a price on it and the prices are randomly dispersed and the average price on all those 100 balls is $100, and you randomly draw 10 of those balls and take the average price on those 10 that sample average will probably not be exactly $100. You draw another 10 and the average on that second sample probably also is not $100 and it probably also doesn't exactly equal the sample average price on your previous sample of 10. This randomness in the prices "drawn" each quarter (the properties sold each quarter by NCREIF members) causes some randomness in the NTBI price LEVEL index. When you compute returns from a price series that has that kind of randomness in it, the RETURNs have negative 1st-order autocorrelation. (Think of it like this... If that first sample you drew happened to have a rather HIGH average, say, $110, then it is more likely that the next sample you draw will have an average BELOW the previous one, and vice versa.)
The negative autocorrelation in the NTBI is not very large, and so that indicates that there isn't too much noise in the NTBI. Noise tends to be worse when samples are small, which happened in the early years of the NTBI and during the financial crisis. Visually, it seems to me the index is noisier in the early years, although statistically the autocorrelation seems about the same.
IF you compute 1st-order autocorrelation at the annual frequency (e.g., on calendar year returns) then you get some slight positive autocorrelation, which is what we would expect in a private market based price index, as prices in the private market do have some degree of inertia. Unlike true volatility, noise does not accumulate over time in the value levels, and hence is diminished (relatively speaking) the lower the frequency of the returns series (e.g., annual vs. quarterly).
Click here for a detailed explanation and example.
Would you like to become a member or subscribe?
Copyright 2014 National Council of Real Estate Investment Fiduciaries
Privacy Statement | Sitemap
Redistribution of any NCREIF data is not permitted without express written consent from NCREIF. Please contact the NCREIF office for more information regarding data usage and redistribution at 312-819-5890.