Friday, August 23, 2013

Tax swapping

The bond sell off of the last few months has created some mark to market losses that may be suitable for some tax loss harvesting, but there is still a portfolio need for income, although not explicitly needed to be tax free. Based on research I've reviewed it feels like there is still relative value in municipal especially compared to other sectors of US fixed income. Some reasons for sticking with muni exposure (for income) include * Detroit headline fears - compared to the Meredith Whitney scare of a few years ago, the Detroit headlines have not caused as severe of a selloff, but they are pretty strong. There are a handful of troubled issuers (Puerto Rico, Chicago, many Michigan cities, etc.) but overall the broader muni obligors are ever so slowly getting better, despite ratings downgrades being slightly tilted to the downside. * Munis relative to duration matched Treasuries are offering more value, given the tax advantages, especially if one is in higher tax state and looks for local funds. Even if one sticks with national funds high tax states like NY/CA tend to be 25-30% of most national funds holdings. Muni's relative to strong corporates are positioned the same way. * If one sticks to the shorter end -- noting that munis are a long duration sector -- there is enough steepness to the curve that perhaps some gains can be picked up from rolldown. * There is tremendous fear of rate rises, but if they do not happen on schedule, there can be money to be had from wearing some of this duration risk. * There is some risk to this, but if one believes that short rates may continue to stay pinned even if tapering starts happening, then levered muni CEF's, trading at discounts, may be particularly attractive as swap vehicles. There are 200+ muni CEF's available. I will filter the candidate universe based on these screens * Current market distribution yield + Discount to NAV greater than 13 (which is the average) This provides some measure of safety (from the discount, as well as some yield) * Expense ratio of less than 110 bps. The Bloomberg is not reliable for this, so this will have to reconfirmed a couple of different ways, both from run rates (on CEFConnect, as well as from the actual filings) Again wide dispersion here, and given that its tough to add value here, why pay up? * Market cap greater than 100mm. Probably colinear with the ER screen, but there is enough illiquidity in CEF space that prowling the very bottom is not needed. * A filter on 90 day volatility, which is a reasonable proxy for duration given the hit that bonds have taken in the last few months. We'll exclude vehicles whose NAV volatility was greater than 9% annualized over the last 90 days. These filters result in a small subset (about 5%) of the vehicles: AFB, MQT, EOT, VKI, VKQ, VPV, IQI, VGM, PMO, MHF Right off the bat, we'll kick out VPV as its a state specific vehicle that is not germane to our needs. Ideally I'm looking for a portfolio with a duration near 7-8 that is investment grade and avoids Puerto Rico, Illinois, Michigan cities, Chicago, Hawaii, and Guam. In addition I'd like dividend coverage of the fund to be high, and I'd like to avoid a situation where there is a high average bond price combined with call exposure. In addition I'd like positive UNII to provide a buffer for div cuts -- which removes MHF.