Giliberto Levy Commercial Mortgage Performance Index Monitor 4Q 2015 Results and Analysis

Transcription

1 Giliberto Levy Commercial Mortgage Performance Index Monitor 4Q 2015 Results and Analysis Table of Contents Market Overview.1 Pricing Trends: 4Q Credit Effects..3 Relative Performance...3 Lending Activity...4 U.S. Commercial Mortgage Outlook.. 5 Appendix: Supporting Tables and Charts 4Q Market Overview The Giliberto Levy Commercial Mortgage Performance Index ended 2015 with a 4Q total return of 0.32%. This was considerably lower than 3Q s 1.57%. The change mainly reflects higher Treasury yields, which pushed most fixed income performance measures down. Fourth quarter total return consisted of 1.18% income, 0.82% price change and 0.04% from other factors. For calendar year 2015, the Index generated 2.70%. This result was well below 2014 s 7.39%. What happened? Giliberto Levy Commercial Mortgage Performance Index Fourth Quarter 2015 Summary Total Return: Total (all sectors) 0.32 (in %) Office 0.45 Multi family 0.19 Retail 0.26 Industrial 0.50 Spreads: Office 185 (as of 12/31/2015) Multi family 175 Retail 195 Industrial 185 Index Statistics: Capitalization ($ billion) (as of 12/31/2015) Duration (years) 4.8 Coupon (%) 4.9 Maturity (years) 7.3 Book LTV (%) 45.8 Spreads are averages for new ten year loans at 60% 65% loan to value ratio. Source: Giliberto Levy Basically capital market winds reversed direction. A salutary combination of (unexpectedly) falling Treasury yields and lower mortgage spreads propelled 2014 s strong results. For example, over the course of 2014, the ten year Treasury yield, measured at month end, declined from 3.04% (December 31, 2013) to 2.17% (December 31, 2014), and the G L composite spread came down from 192 basis points (bp) to 158 bp. That produced strong markto market capital value increases, complementing coupon yield. In 2015, the opposite occurred. The ten year yield started at the aforementioned 2.17% and ended at 2.27%. What really cut, though, was an increase in credit spreads, which ticked up to 185 bp in September and remained at that level through year end. With the Fed finally putting in place a long expected modest hike in short rates in December 2015, Treasury yields at all maturities along the yield curve moved up (see Charts 1 and 2). The benchmark ten year yield added 21 bp. Key yields from six months to seven years were up 30 to 40 bp; increases at the long end were smaller. As a result, loans with maturities below ten years posted larger price declines than long term assets, despite the shorter durations of the former group. 1The Index's components are fixed rate commercial mortgage loans held on balance sheets of institutions such as life insurance companies and pension funds. Index returns are a market valueweighted blend of office, apartment, retail, industrial, lodging, mixed use and other miscellaneous property types. Index performance tracks senior loans only; it does not include construction loans, mezzanine and other subordinate instruments and bridge loans made by such institutions. Introduction The Giliberto Levy Monitor reports the quarterly results of the Giliberto Levy Commercial Mortgage Performance Index SM and provides comprehensive indepth market analysis and commentary on key aspects of the commercial mortgage industry. Income, price, total returns and spreads are provided for office, apartment, retail, industrial and other property after adjusting for credit loss. Additional Index statistics include capitalization, duration, coupon rate, maturity and loan to values, and comparison of relevant returns and spreads to other debt classes such as investment grade and U.S. Treasury bonds. A subscription to the Giliberto Levy Monitor includes access to the Giliberto Levy Analyzer a custom query tool allowing users to analyze total return by property on a long term historical basis. 4Q 1

2 Our core property sector composite spread was unchanged over its September 30 level, remaining at 185 bp (see Chart 3 and Table 1). 2 This contrasted with an approximately 15 bp upward shift in spreads that occurred for investment grade CMBS, according to Barclays data. As a result, CMBS underperformed the G L. Chart 1. US Treasury Yields, September 30, 2015 and December 31, 2015 (in %) Pricing Trends: 4Q 2015 Our composite credit spread, as noted previously, was essentially unchanged. We recorded a 5 bp decline in apartment spreads. Other major sectors showed insignificant increases; lodging and mixed use spreads also showed little change. Miscellaneous others (parking, storage, etc.) increased 5 bp to 10 bp, but that may reflect differences in the composition of collateral types aggregated into that grouping. (Our standardized spreads take into account differences in factors such as term, LTV, etc.) We measured a substantial decrease in the mortgage term premium for three year deals: It declined from 100 bp to 55 bp. Interestingly, this 45 bp decline almost exactly offset the 39 bp increase posted by the three year Treasury yield during the quarter. The five year term premium stayed at 30 bp. Although the five year term premium was unchanged, new five year loans lost some of their yield advantage over tenyear deals: As of quarter end the edge was down to 21 bp from 3Q s 39 bp. In general, seven year deals did not show a term premium. Sources: U.S. Treasury; Giliberto Levy Chart 2. Changes in US Treasury Yields, September 30, 2015 to December 31, 2015 (in basis points) Table 1. Key Rates Key Rates (in %) 9/30/ /31/2015 change 20-year Treasury Yield year Mortgage Spread Year Mortgage Yield year Treasury Yield year Mortgage Spread Year Mortgage Yield Sources: U.S. Treasury; Giliberto Levy Chart 3. Spreads over Treasuries for Ten year Commercial Mortgages (in basis points, reflecting 30/360 day count convention) 7-year Treasury Yield year Mortgage Spread Year Mortgage Yield year Treasury Yield year Mortgage Spread Year Mortgage Yield year Treasury Yield year Mortgage Spread Year Mortgage Yield Sources: U.S. Treasury; Giliberto-Levy 2 The composite measures ten year loans with 60% to 65% loan to value (LTV) ratios. 4Q 2 Source: Giliberto Levy. Chart data are averages of office, apartment, retail and industrial property loans.

3 The average LTV on new loans eligible for inclusion in the Index this quarter was 62%, not meaningfully higher than last quarter s 60%. Average DSC dropped to 1.97 times from the prior quarter s The decline was driven by lower DSCs for office, retail and other. (Please note that data for 3Q reflect slight revisions from previously reported values due to September 2015 loans reported late.) Ten years continues to be the most popular maturity. For these loans, most LTVs fell in the 50% 80% range. Segmenting that range, we measured median spreads of 188 bp for 50% 60% LTV, 200 bp for 60% 70% and 209 bp for 70% 80%. However, the overall dispersion of spreads for 50% 80% LTV ten year loans was quite wide: at the low end, around 100 bp and at the high end, slightly over 300 bp. (Unlike our standardized spreads, these median spreads by LTV range are not adjusted for differences in loan size, prepayment type and so forth.) We found some movement in loan size premiums. For most sectors, loans for $5 million and lower carried 20 to 25 bp premiums; loans between $5 million and $10 million had 10 to 15 bp premiums. Apartments were an exception: There was no apparent differentiation among loans smaller than $10 million. A 15 bp premium appeared for these loans. Importantly, size premiums are measured averages. They did not universally appear in all small balance loans. Last quarter, all sector spreads increased, but apartment spreads increased the least. Apartments also have the longest average duration. As a result, the apartment sector benefitted the most from favorable (absolute) declines in Treasury yields and favorable (relative) changes to credit spreads. It was different in 4Q. Apartments still had a favorable spread change, but it was small. However, the sector s long duration caused values to be whacked by the rise in Treasury rates. Apartment returns were also pulled down relative to the overall Index by the sector s lower income yields. Importantly, the lower income return does not reflect credit impairment far from it! Lower income yield is in place due to lenders generally accepting lower yields in the sector. Index results, while modest, contrasted starkly with the negative returns posted by most comparable investmentgrade fixed income sectors (see Table 2). The Index also ran well ahead of other segments for calendar year 2015 results. Chart 4. Commercial Mortgage Credit Effects, 005 to 4Q 2015 (Rolling four quarter effect as a percentage of book value, in basis points) Credit Effects Credit effects remain a non factor in Index returns at this point in the cycle. In fact, our credit measure hit an alltime historic low, posting a 5 bp trailing four quarter run rate as of year end 2015 (see Chart 4). Relative Performance As usual, most of the variation in sector returns this quarter is accounted for by variation in price return. Price return differences are driven by relative shifts in yield changes and differences in sector duration. When yields decrease, the sector with the longest duration gets the biggest price appreciation, all else equal. Sources: Giliberto Levy Table 2. Total Return as of December 31, 2015 Total Return (%) Duration Index 4Q 2015 YTD Last 4 Q as of 12/31/2015 Commercial Mortgages* Duration-adjusted Baa Bond CMBS: Investment Grade CMBS TSY Int. Term TSY 5-7 yr. Term Credit Int. Term A Int. Term Baa Int. Term High-yield Int. Term Ba Int. Term Q 3 * Giliberto-Levy Total Index, includes office, retail, apartment and industrial and other Sources: Barclays; Giliberto-Levy

4 It s instructive to focus on the duration adjusted Baa bond return to understand why the Index outperformed. First, its income return in 2015 was about 72 bp higher. That leaves about 185 bp attributable to differences in price movements and other factors. The big differentiator is yield change. The Index s weighted average yield moved up 27 bp, from 3.34% at year end 2014 to 3.61% at year end In contrast, the yield on Baa intermediate bonds increased63bpperbarclays.durationaveragedacrossthe year was roughly 4.6. If we multiply 4.6 (duration) times 36 bp (the difference between 27 bp and 63 bp), we get about 1.7%, which accounts for most of the performance gap not explained by income return differences. New loans totaling more than $8.0 billion were added to the Index (see Table 3 for details). This came from loans originated in September, October and November December 2015 loans, which are included in the production number cited above and depicted in Chart 5, will be added to the Index for 016. Chart 5. Index eligible Loan Production, 009 to 4Q 2015 Lending Activity 4Q origination of loans eligible for the Index came in at roughly $6.6 billion, compared with 3Q 2015's (revised) $9.1 billion (see Chart 5). This apparent shortfall will decrease as we receive information on loans made during December 2015 that have not yet been reported. For the first three quarters of 2015, the difference between initial and revised production averaged 24%, so this gap can be substantial. Also, these revisions do not include latereported loans that occurred prior to the final month of the previous quarter. For example, we received reports of some August 2015 loans during our 4Q 2015 data gathering. Those and similar 2015 late reports will enter the Index in its 016 release. Full term interest only loans accounted for 12% of total lending volume in our data, a meaningful drop from 3Q s 23%. Not surprisingly, such loans had, on average, lower LTV and higher DSC. Loans with full amortization also declined to about a 6% share, from 9% in the prior quarter. Loans that partially amortize over their terms make up the remainder. About 72% of these loans, as measured by volume, had interest only periods. That s a large jump from last quarter s 43%. This increase propelled the percentage of loans with either full or partial interest only periods to almost 71% of total volume for the quarter, compared with about 53% in 3Q. Statistics are subject to revision as we receive additional data on 4Q 2015 production. Sources: Giliberto Levy Table 3. New Loan Profile, 4Q 2015 Profile of Loans Added to Index Fourth Quarter 2015 Weighted-Average Amount Number Coupon Spread DSCR LTV Term ($) of Loans (bp) (%) (years) Office 1,546,079, % Apartment 2,320,210, % Retail 1,867,290, % Industrial 1,536,650, % Other 764,000, % Total 8,034,230, % Source: Giliberto-Levy Please note that spreads reported in Table 3 are weighted averages of loans made during periods prior to December These averages do not represent quarter end pricing parameters. Those pricing spreads are reported in the summary table at the top of this report for standardized tenyear loans at 60% to 65% LTVs. These standardized spreads, with the addition of term premiums, also appear in Table 1. Standardized spreads are generated from new loan originations and form the baseline for marking to market all loans in the Index. 4Q 4

5 US Commercial Mortgage Outlook Last year at this time, we provided an analysis of loans set to mature in The purpose was to understand both the volume of loans and the extent of rollover risk. We decided to repeat this exercise for loans maturing in About $18 billion of Index principal, as measured by December 31, 2015 balances, is set to mature in This accounts for about 10% of outstanding Index balances. (The percentage is biased upward because we have not yet added loans that were made in December 2015 to the Index. This doesn t affect the maturing balance amount because the likelihood of a loan suited for the Index that was made in December 2015 with a maturity of one year or less is near zero.) We divided maturing loans into two groups: (1) those that amortize fully and (2) those that have balloon payments. Based on year end balances, (1) accounts for about 5.9% of total maturing principal. Naturally, loans that have 12 months or less remaining and are fully amortizing have quite low balances relative to collateral value. We d also expect them to have literally no risk of default at maturity. However, some of these loans might seek new financing. We included a row in Table 4 for loans that were missing DSCs at origination. All these loans fell into lower LTV buckets, so we doubt they pose significant risk. Highlighted entries in Table 4 are where loans may have difficulty refinancing in full at maturity. In total, they don t make up even 0.5% of maturing loans with balloons. Last year, this statistic was closer to 1.3%. Keep in mind that all loans included in that 1.3% did mature. The historically low credit event effect registered in 2015 implies few if any issues were encountered. The conclusion is that 2016 is likely to be very similar: extremely low re fi risk and frictional credit effects. The fun in 2016 is more likely to come from gyrations in capital markets. Case in point: As we were finishing up these comments the ten year US Treasury yield was 50 bp lower than where it ended Who predicted that? As always, your comments and suggestions are welcome. February 2016 Since LTV of the now paid off loan will be zero, refinancing concerns will focus on DSC. Only about 1.1% of these loans have current DSCs below 1.15, so nearly all loans that seek refinancing generally should qualify. As you d expect, the story is somewhat more involved for the 94.1% of maturing loans that have a balloon obligation. We sorted these by current LTV and DSC, with results in Table 4. Table 4. Breakdown of Loans with Balloon Payments at Maturity in 2016 Source: Giliberto Levy 4Q 5

6 Note on Monitor Exhibits In 015 we added a new exhibit that uses book LTV rather than market LTV to stratify Index results. The reason for the new page is that market LTV, which uses marked tomarket loan values, has been affected by low interest rates. Values of many loans are now sufficiently above par that the resultant market LTVs may not be indicative of credit risk. To rectify that, we introduced a similar page that uses book value divided by current collateral value. This measure is still partially marked to market: collateral value change is picked up. This ratio of book value, i.e., current loan balance, to current collateral value is, we believe, a better indicator today. Starting with 016, we will no longer present the exhibit that uses market LTV. Chart 6. Commercial Mortgage Performance Rolling four quarter total return for Total index Source: Giliberto Levy Chart 7. Commercial Mortgage Credit Effects Rolling four quarter losses as a percentage of book value, in basis points Source: Giliberto Levy 4Q 6

7 Appendix: Supporting Tables and Charts Q Tables and Charts of Results A Giliberto Levy Performance by Sector 9 B Giliberto Levy Statistics & Averages 9 C Giliberto Levy Credit Quality Using Market Value 10 D Giliberto Levy Credit Quality Using Book Value 11 E Giliberto Levy Coupon Rate 12 F Giliberto Levy Remaining Term 13 G Giliberto Levy Duration Cell 14 H Giliberto Levy Vintage Cohort 15 4Q 8

8 Table A. Giliberto Levy Index by Sector Returns (%) Credit Losses (book value; bp) Last 3 months YTD Last 12 Index Last 3 Last 12 Sector Income Price Other Total Total Months Level Months YTD Months Office Apartment Retail Industrial Other Sectors N/A Total Aggregate Aggregate excludes Other Sectors (hotel/motel, mixed-use and miscellaneous) Other Return is paydown return and compounding Index bases: Aggregate Dec = 100; Total and major sectors Dec = 100 bp basis points Table B. Giliberto Levy Total Index Sector Statistics Returns (%) Statistics Averages Adj. Pct. of Num. of Sector Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts Office % 3064 Apartment % 3401 Retail % 4448 Industrial % 2882 Others % 1555 Total % Aggregate % Aggregate excludes Other Sectors (hotel/motel, mixed-use and miscellaneous) 4Q 9

9 Table C. Giliberto Levy Total Index by Credit Quality Using Market Value Returns Statistics Averages Credit Quality Adj. Pct. of Num. of Using Market Value Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts Investment-grade % Crossover % 260 High-yield % 2 Total % Credit quality definition Marked to Market LTV Minimum Maximum Investment-grade 0.0% 70.0% Crossover 70.1% 85.0% High-yield 85.1% NA 4Q 10

10 Table D. Giliberto Levy Total Index by Credit Quality Using Book Value Returns Statistics Averages Credit Quality Adj. Pct. of Num. of Using Book Value Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts Investment-grade % Crossover % 259 High-yield % 2 Total % Credit quality definition Book LTV Minimum Maximum Investment-grade 0.0% 70.0% Crossover 70.1% 85.0% High-yield 85.1% NA 4Q 11

11 Table E. Giliberto Levy Total Index by Coupon Rate Returns Statistics Averages Adj. Pct. of Num. of Coupon Rate Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts 0.0% to 4.0% % % to 4.5% % % to 5.0% % % to 5.5% % % to 6.0% % % to 6.5% % % to 7.0% % % to 7.5% % % to 8.0% % % and above % 994 Total % Q 12

12 Table F. Giliberto Levy Total Index by Remaining Term Returns Statistics Averages Adj. Pct. of Num. of Maturity Cell Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts 0 to 2 years % to 4 years % to 6 years % to 8 years % to 10 years % to 12 years % to 14 years % to 16 years % to 18 years % to 20 years % years or more % 366 Total % Q 13

13 Table G. Giliberto Levy Total Index by Duration Cell Returns Statistics Averages Adj. Pct. of Num. of Duration Cell Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts 0 to 1 year % to 2 years % to 3 years % to 4 years % to 5 years % to 6 years % to 7 years % to 8 years % to 9 years % to 10 years % years or more % 208 Total % Q 14

14 Table H. Giliberto Levy Total Index by Vintage Year Returns Statistics Averages Adj. Pct. of Num. of Vintage Year Income Price Other Total Duration Coupon Maturity Price Yield Portfolio Cohorts before % % % % % % % % % % % % % % % % % 1311 Total % Q 15

15 About Giliberto Levy Index Established in 1993, the Giliberto Levy Index provides essential information and accurate research on emerging trends affecting commercial mortgage investments. Investment professionals, actuaries, consultants, plan sponsors, banks, life companies, regulators and rating agencies use the index as a reliable way to evaluate the performance of commercial mortgages on a total return, marked to market basis. Disclaimer 2016 John B. Levy & Company, Inc. All rights reserved. No part of the Giliberto Levy Commercial Mortgage Performance Index (GLCMPI) may be reproduced or transmitted, in any form or by any means, without the prior written consent of John B. Levy & Company, Inc. This index is neither appropriate nor authorized by John B. Levy & Company, Inc. for use as a benchmark for portfolio or manager performance, or as the basis for any business decision. John B. Levy & Company, Inc. gives no warranty or representation that the use of this information will achieve any particular result for you. John B. Levy & Company, Inc. has no liability for any losses, damages, costs or expenses suffered by any person as a result of any reliance on this information. Contributors The Giliberto Levy Index pulls specific data from several sources to assure the accuracy of its measurement. Information on mortgage rates comes from the John B. Levy & Company National Mortgage Survey. Data on interest rates come from market sources. Contact To subscribe or receive more information on the Giliberto Levy Commercial Mortgage Performance Index SM Monitor, please contact jlevy@jblevyco.com. John B. Levy & Company, Inc. 830 East Main Street, Suite 2100 Richmond, Virginia Q 16