Thank you, Carlos, and good morning. Despite the continued challenging economic and fiscal conditions in our main market, overall asset quality remains stable during the second quarter of the year.
In Puerto Rico, credit quality metrics reflect lower NPLs, lower NPAs, lower NPL inflows and higher net charge-offs. In the U.S, credit metrics reflect strong portfolio growth, stable NPLs, stable NPL inflows and stable net charge-offs.
Please turn to Slide #5. As Ignacio covered earlier, our current outstanding current exposure to the Puerto Rico government, municipalities and other instrumentalities is $517 million, relatively unchanged compared to the prior quarter. After quarter end, we sold the remaining central government exposure.
Our municipality exposure consists mainly of senior priority loans to a select group of municipalities, whose revenues are independent of the central government. In most cases, the good faith, credit and unlimited taxing power of each municipality is pledged to the repayment of the loans. Our top 4 exposures are to Carolina, where the airport and several major tourist hotels are located; San Juan, the capital of Puerto Rico; Guaynabo, the municipality with the highest per capita income; and Bayamón, the second most populous municipalities. These municipalities comprise 77% of our total exposure.
As discussed in previous webcasts, the oversight board certified an amended 10-year fiscal plan for Puerto Rico in February. Among other things, the fiscal plan provides for reduction of general fund subsidies to Puerto Rico's municipalities. Such subsidies constitute a material portion of the operating revenues of certain smaller municipalities, though only representing 7% of revenues for the largest municipalities that make up our portfolio.
We also have indirect lending facilities in which the government acts as a guarantor. The largest such exposure is in the form of residential mortgage loans to individual borrowers in which the government provides a guarantee similar to FHA programs in the U.S.
Turn to Slide #6 to discuss credit metrics for the quarter. Nonperforming assets, including covered loans, decreased by $38 million from $795 million in the previous quarter to $758 million this quarter. The decrease in nonperforming assets was the combination of a decrease of $28 million in nonperforming loans and a decrease of $9 million in OREOs. At the end of the quarter, the ratio of nonperforming assets to total assets stood at 1.8%, decreasing by 14 basis points from the previous quarter. Nonperforming loans in Puerto Rico decreased by $31 million from the first quarter of 2017, driven by lower commercial and mortgage NPLs of $12.4 million and $12.2 million, respectively, mostly due to net charge-off activity.
In the U.S., nonperforming loans increased by $3 million, mainly due to the consumer portfolio. The ratio of NPLs to total loans remain stable at 2.4% from 2.5% in the first quarter of 2017. The decrease in OREOs was mainly driven by the Puerto Rico mortgage portfolio.
Please turn to Slide 7 for a summary of the trends in NPL inflows. Compared to the prior quarter, NPL inflows, excluding consumer loans, decreased by $20 million, mostly in Puerto Rico. The decrease in Puerto Rico was mainly driven by the inflow in the previous quarter of a single commercial relationship of $25 million. Excluding this relationship, inflows to NPL were relatively flat quarter-over-quarter. In the U.S., NPL inflows were flat at $6 million.
Please turn to Slide #8. Net charge-offs for the quarter amounted to $57 million or annualized 101 basis points of average loans held-in-portfolio, compared to $36 million or 63 basis points in the prior quarter. The $21 million increase in net charge-off was mainly in Puerto Rico as net charge-off in the U.S. remains stable at $3.2 million. The increase in net charge-off in Puerto Rico was mainly related to a $9 million increase in commercial portfolio, mostly due to a $12 million charge-off taken on a previously reserved single relationship and a $7 million increase in the mortgage portfolio. The increase in mortgage net charge-off was mostly due to a one-time $6 million impact associated with residential mortgages related to a residential project acquired from Doral that suffered significant structural damages.
The provision for loan losses, including covered loans, increased by $12 million from the prior quarter driven by an increase in charge-off activity during the quarter. The ratio of provision to net charge-off stood at 87% during the second quarter compared to 118% in the previous quarter. The allowance for loan losses decreased by $8 million from the first quarter to $509 million, mainly due to a decrease of $12 million in Puerto Rico, offset in part by the increase of $5 million in the U.S. The Puerto Rico decrease was mainly attributed to the aforementioned $12 million previously reserved single commercial relationship.
The U.S. increase was mainly driven by higher reserves for the taxi medallion portfolio. At the end of the second quarter, our taxi medallion portfolio acquired in the Doral acquisition had an unpaid principal balance of $235 million net of reserves. The current value of this portfolio is $130 million or approximately 55% of its unpaid principal balance, representing less than 1% of our total loan portfolio. 96% of the portfolio is in New York City, with an average carrying loan value of $364,000 per medallion. The ratio of allowance to loan losses -- to loans held in portfolio remained flat at 2.3% quarter-over-quarter. The range of the allowance for loan losses to NPLs increased slightly to 93% compared to 90% in the previous quarter.
To summarize, despite challenging economic conditions in our main market, our credit metrics for the second quarter of the year remains stable. The continued improvement in our credit risk profile is the result of the steps we have taken to derisk our loan portfolios by reducing exposure to high-risk asset classes and the improvements in our credit underwriting criteria. Given uncertainties in our main market, we remain attentive to changes in credit quality trends.
With that, I would like to turn the call over to Ignacio for his concluding remarks. Thank you.