Private sector lender Yes Bank has an overall stressed advance of 8 percent and new chief executive Ravneet Gill's clean-up will keep profit under pressure for up to the next 18 months, rating agency Moody's warned Tuesday.

However, looking beyond the near-term stress, Gill's clean-up is positive, once the de-risking is completed, the agency noted. "We estimate that Yes Bank's overall stressed assets are about 8 percent of its total loans, taking into account this new disclosure," Moody's said after the bank Friday reported its maiden loss of Rs 1,506 crore for March quarter.

Gross non-performing assets shot up to 3.2 percent, while Gill also marked out an Rs 10,000-crore portfolio as potentially stressed. Provisions on account of higher NPAs and also a contingent provision representing 20 percent of the Rs 10,000 -crore portfolio resulted in the losses, Gill explained.

The agency said the bank expects 50 percent of these potentially stressed loans to slip into NPAs and will provide more for the same going forward. The 8 percent dud loans include a gross NPAs of 4 percent which has been called out as potentially stressed, net standard restructured loans and security receipts of 0.8 percent of gross loans, the agency said.

On a pro-forma basis, the provision coverage ratio is at 33 percent of the stressed assets, it added. "The balance sheet clean-up will strain the bank's profitability in the next 12-18 months as it provides for these higher quantum of stressed assets," Moody's warned.

The agency also said the shift away from corporate loans to retail and small businesses will also strain the bottom line as the corporate loans have been a key source of earnings. But the shift is credit positive as such loans are better from an asset quality perspective, it added.

The agency further noted that the bank has decided to slow down on loan growth to 20-25 percent, as against 34 percent average of recent years. It can be recalled that on Monday in an unusual step, Australian brokerage Macquarie made a public apology saying they had got their call on Yes Bank all wrong and forecast a 40 percent correction in stock price and downgraded it by two notches.

Admitting to overlooking the risks from the structured finance business of Yes Bank, Macquarie said, "we must eat the humble pie today and admit we underestimated the risks in structured finance. We got the call wrong."

The brokerage also flagged concerns on the fee income and the retail franchise of the fifth largest private sector lender. It also said over the past eight years, it felt the bank can thrive in a risky business like structured finance.

The bank led by Rana Kapoor, who was forced out by the Reserve Bank of India (RBI) earlier this year for reportedly under-reporting bad loans and for poor governance standards, reported a three-times increase in BB-rated and below accounts despite a higher slippage of Rs 3,408 crore in the quarter is a negative surprise.

The bank made higher provisions for possible reverses, including a massive Rs 2,100-crore contingency reserves, leading to the massive loss and said had it not been for an Rs 831-crore write-back, it would have reported higher losses for the March quarter.

It also saw an almost ten-times spike in provisions to Rs 3,661 crore from Rs 399 crore in the year-ago period. This includes a contingent provision of Rs 2,100 crore on an Rs 10,000-crore exposure to potentially stressed assets in real estate, media & entertainment and infra sectors, which stands at risk, the management told analysts on a concall.

Overall slippages jumped to Rs 3,481 crore, including Rs 552 crore to Jet Airways and Rs 529 crore for the bankrupt infra lender IL&FS. The gross NPA ratio more than doubled to 3.22 from 1.28 in the year-ago period and 2.10 in the preceding quarter. It has a 7 percent exposure to the commercial realty sector, which is facing troubles.

There was a bloodbath on the stock as investors punted the Yes Bank counter, which plunged close to 30 percent to Rs 168 on the BSE after hitting a low of Rs 165.30 against a flat closing of the benchmark.