订阅以获取KKR的最新洞察
自动质量提示本研报仍可阅读，但 AI 分析或译文低于优选质量阈值，已进入改进队列。重要判断请同时核对官网原文。
完整研报正文
完整中文译文
高收益债的第二幕：AI揭示了信用质量
要点
注册订阅KKR的最新见解
结构性质量提升、行业集中度下降以及发行人需求增长——这些因素结合在一起，正是我们所说的“高收益债可能正在进入其第二幕”。
高收益债可以说是1980年代最早的新型融资渠道之一，并且一直是杠杆融资生态系统的核心部分，但其增长一度停滞——过去十年年均增长率仅为0.9%。近期发生的变化在于，结构性质量改善、发行人偏好演变以及贷款和直接贷款技术面的收紧，这些因素正同时汇聚。
质量故事，而非仅仅是技术性故事
高收益市场在过去近十年里悄然演变成与许多投资者记忆中的资产类别不同的存在。源自不同时代的“垃圾债券”标签已不再符合该市场如今所代表的内涵。美国及欧盟高收益市场目前以创纪录的约57%和68%的BB级占比，其信用质量更接近投资级领域，而非信用光谱中更具投机性的一端。这一演变并非一蹴而就——它反映了数十年的市场发展和发行人成熟，而新冠疫情时期的“堕落天使”则催化了部分最高质量的杠杆借款人逐步迁移至高收益债券市场。
高收益指数与杠杆贷款市场的构成对比具有启发性。尽管贷款市场约60%为单B级，高收益市场的BB级群体仍在持续扩大。在美国，BB级部分已从2026年1月的$1.19万亿美元增长至5月的$1.25万亿美元，而CCC级部分在同一时期从$213亿美元缩减至$193亿美元。这一分化之所以重要，是因为它意味着这两个市场代表了显著不同的风险特征，而随着杠杆融资领域的离散度扩大，这种区别变得更加重要。
图表1：美国高收益债按优先级分类的占比
另外两个质量维度常常被忽视。首先是久期。大约3年的有效久期接近15年来的低点，远低于约4年的长期平均水平。这一点很重要，因为尽管利差看似紧张（接近全球金融危机以来的96百分位），但考虑较低的久期后，情况则有所不同。每单位久期的利差仅处于第74百分位，这比表面数字所显示的状况要宽松得多。其次是结构性优先级别。优先留置权担保债券现在占市场的33%，创历史新高，在美国和欧盟的总担保敞口分别接近36%和38%。美国数据中心融资市场的增长，目前约占指数的3%，且几乎全部为担保债务，是这一趋势的近期推动因素之一。这使我们更加确信，高收益市场的构成受到高质量债券和结构的支撑。
今年早些时候的软件板块抛售凸显了这一点。软件板块约占美国和欧洲高收益市场的3%，而美国杠杆贷款市场约13%，美国直接贷款市场超过20%。当人工智能驱动的颠覆担忧引发软件信贷的广泛重新定价时，影响与敞口成正比。软件贷款的买价跌至四年低点，软件和非软件贷款买价之间的差距在Q1末扩大到超过8个百分点。高收益债由于其结构性较低的软件集中度，表现明显更好。该资产类别并未逃脱审查，但它以不同的方式吸收了冲击，我们认为这一差异值得深入探讨。
质量故事并未止步于信用评级，还延伸到报告和披露。由于高收益债券根据证券法注册（即144a注册），发行人通常遵守更高的披露标准，包括发行备忘录（OM）、季度财报电话会议，以及许多情况下完整的10-K和10-Q文件。
图表 2：美国高收益债券行业构成
融资方式的转变
从历史上看，高收益债并非许多发起人的首选。赎回保护的经济性、SEC注册要求带来的额外流程，以及对灵活浮动利率结构的偏好，使得杠杆贷款以及最近的直接贷款成为大多数发起人支持交易的首选融资渠道。
随着CLO的需求变得更加挑剔，直接贷款的条款和风险收紧，债券市场提供的执行确定性变得更加珍贵。不可赎回的高收益债券在特定时期内锁定固定资本成本，消除了重新定价风险，并触及多元化的投资者群体，这些投资者不受近年影响贷款市场技术面的CLO需求动态的制约。对于现金流稳定且预期利率最终将向有利方向发展的发行人来说，支付略高的票息以换取结构确定性和赎回保护，正成为越来越有吸引力的交易。
近期的交易活动反映了这一转变。正如我们在最近的报告《超越喧嚣》中强调的，美国高收益债券发行量在4月达到了七个月高点，而BB级债券占年初至今新发行的2026,，创下任何可比时期的历史新高。此后，数据持续朝同一方向发展。截至7月初，年初至今的预计发行量为39%亿美元，比去年同期的$189亿美元高出超过20%。在欧洲，高收益BB级债券的过去十二个月发行量接近今年早些时候创下的历史高位。而发行构成则讲述了更有趣的故事。BB级债券在6月至今的供应量中占比达到$157，远高于5月的59%，而单B级发行量则骤降至不足30%，创下9%个月以来的新低。
以往完全通过贷款市场融资的交易，现在越来越需要债券市场的参与才能完成，EA Sports和Sealed Air就是近期的典型例子：贷款市场无法独立消化交易，而债券市场的参与完成了这笔交易。我们认为这并非孤立事件，而是结构性变化的信号。事实上，贷款市场约有15的债务在两年内到期，这意味着贷款市场中超过7%亿美元的面值需要考虑高收益债是否是个更好的选择。此外，高收益市场也有约$100的债务在未来两年内到期，这是自金融危机前以来的最高比例。
尽管新交易日益依赖高收益市场，我们认为即将到来的再融资可能会更显著地纳入高收益债券。Asurion提供了一个具有启发性的例子。就在去年，Asurion还是一家仅发行贷款的发行人，未偿贷款约$12亿美元，分属六个批次。面对2026至2028,年间的近期到期期限，且许多现有投资者已对该名称持仓饱和，公司需要新的策略。通过转向高收益市场，公司接触到了不同的投资者群体，并于今年早些时候定价发行了$3.3亿美元的有担保债券，用以偿还现有批次。偿还债务恢复了贷款市场的信心，使剩余批次利差收紧，并降低了后续再融资成本。随之而来的良性循环并非偶然，而是开辟新融资渠道的直接结果。展望未来，我们预计许多软件发行人，尤其是那些展现出强劲现金流生成能力和稳定盈利能力的公司，将走类似的路径。高收益市场不仅提供了替代选择，在某些情况下，甚至是更优的途径。
这对投资者意味着什么？
如今高收益债券的投资逻辑并非押注利差收窄，而是一种结构性再配置论点，具有多种回报路径。对于在两个市场建立敞口的配置者而言，美国和欧洲的机会集并非同步变动——行业构成、发行人基础以及利率周期时点的差异足以使得两者相结合能带来单一市场无法提供的多元化收益。
首先是收益率本身。目前ICE BoA高收益指数约为~7.3%，在固定利率确定性愈发宝贵的当下，该资产类别提供了具有吸引力的绝对收益。在欧洲，ICE BoA欧洲高收益当前收益率约为~5.5%，而进行外汇对冲的投资者还可额外获得约~150 basis points的利差。一只锁定五至八年票息且附有赎回保护的债券，与在波动利率环境中每季度重新定价的浮动利率贷款是截然不同的选择。
值得注意的是，美国Morningstar LSTA杠杆贷款指数收益率目前为8.1%，比高收益债券指数高出约~100 basis points。对于愿意提升信用质量的投资者而言，这一溢价已不再反映更优的基本面，而是反映了贷款市场中积累的软件积压、CLO技术性压力以及文件问题。换言之，贷款市场如今提供更高收益率恰恰是因为其承担了更多风险。高收益债券在当前水平可能是更具吸引力的风险调整后入场点。
其次是结构中所蕴含的期权性。当发行人提前再融资时——正如在信贷获取改善时期一贯发生的那样——以折价购买的投资者获得的回报远高于按最差收益率计算的模型值。当债券未被赎回且仍以高于面值流通时，票息持续复利，实际回报往往超过初始预期。当并购活动加速时，以折价持有的债券按面值被赎回，产生的内部收益率是单纯利差收窄所无法实现的。
对于一直低配高收益或多年未重新审视该主题的配置者而言，问题很简单：投资组合反映的是该市场的真实现状，还是其过去的形象？结构性质量正在提升，发行人需求不断增长，每单位风险所取得的绝对回报颇具吸引力。第一幕耗时十年方才上演。第二幕已经开始。
本文作者Christopher Sheldon，发表于3,年8月2026日《金融时报》。
完整英文原文
Sign up to subscribe to the latest insights from KKR
Key Takeaways
Sign up to subscribe to the latest insights from KKR
The combination of structural quality, reduced sector concentration, and growing issuer demand is what we mean when we say high yield may be entering its second act.
High yield was arguably one of the first novel financing channels of the 1980s and has remained a core part of the leveraged finance ecosystem, but its growth had stalled -- a mere 0.9% annualized over the past decade. What has changed recently is that structural quality improvements, evolving issuer preferences, and tightening loan and direct lending technicals are all converging at once.
A Quality Story, Not Just a Technical One
The high yield market has spent the better part of a decade quietly becoming something different from the asset class many investors remember. The “junk bond” label, inherited from a different era, has not kept pace with what the market represents today. At a record ~57% and 68% BB-rated composition, the U.S. and EU high yield market now sits closer in quality to the investment grade universe than to the more speculative end of the credit spectrum. That evolution did not happen overnight – it reflects decades of market development and issuer maturation, with COVID-era fallen-angels catalyzing a gradual migration of some of the highest-quality leveraged borrowers into the high yield bond market.
The composition contrast of the high yield index with the leveraged loan market is instructive. While the loan market is approximately 60% single-B rated, the high yield market has continued to see its BB cohort expand. In the U.S., the BB cohort has grown from $1.19 trillion in January 2026 to $1.25 trillion in May, while the CCC bucket shrunk from $213 billion to $193 billion over the same period. That divergence matters because it means the two markets represent meaningfully different risk profiles, and the distinction has become more consequential as dispersion across leveraged finance has widened.
EXHIBIT 1: % of Total U.S. High Yield by Seniority
Two additional dimensions of quality often go unobserved. The first is duration. At roughly 3 years, the market's effective duration sits near a 15-year low and well below its long-run average of approximately 4 years. That matters because while spreads appear tight (near the 96th percentile since the GFC), adjusting for lower duration tells a different story. The spread per unit of duration sits at only the 74th percentile, a meaningfully less stretched picture than the headline figure suggests. The second is structural seniority. First lien secured bonds now represent 33% of the market, an all-time high, with total secured exposure approaching 36% and 38% in the U.S. and EU respectively. The growth of the data center financing market in the U.S., now roughly 3% of the index and almost entirely secured, has been one of the more recent contributors to that trend. This reinforces to us that the composition of the high yield market is supported by high quality paper and protected by structure.
The software selloff earlier this year crystallized this point. Software represents roughly 3% of the U.S. and European high yield market, compared to approximately 13% of the U.S. leveraged loan market and over 20% in U.S. direct lending. When AI-driven disruption concerns triggered a broad repricing of software credit, the impact was proportional to exposure. Software loan bids fell to the lowest reading in four years, and the gap between software and non-software loan bids widened to more than eight points by end of Q1. High yield, with its structurally lower software concentration, held materially better. The asset class did not escape scrutiny, but it absorbed the shock differently, and we believe that difference is worth double-clicking into.
The quality story doesn’t end at credit ratings. It carries into reporting and disclosure. Because high yield bonds are registered under securities laws (i.e. 144a registration), issuers are generally subject to a higher level of disclosure, including offering memorandums (OMs), quarterly earnings calls, and in many cases full 10-K and 10-Q filings.
EXHIBIT 2: U.S. High Yield by Sector Composition
A Shift in Financing
High yield has not historically been the first call for many sponsors. Call protection economics, the additional process associated with SEC registration requirements, and the preference for flexible floating-rate structures made leveraged loans and, more recently, direct lending the default financing channels for most sponsor-backed transactions.
As CLO appetite has grown more selective and direct lending terms and risk have tightened, the execution certainty that the bond market offers has become more valuable. A non-call high yield bond locks in a fixed cost of capital for a defined period, eliminates repricing risk, and accesses a differentiated investor base that is not subject to the same CLO-driven demand dynamics that have shaped loan market technicals in recent years. For issuers with stable cash flows and a view that rates will eventually move in their favor, paying a modestly higher coupon in exchange for structural certainty and call protection is an increasingly attractive trade.
Recent deal activity reflects this shift. As we highlighted in our recent note Beyond the Roar, U.S. high yield issuance reached a seven-month high in April 2026, and BB-rated paper accounted for 39% of new issuance year-to-date, an all-time high for any comparable period. Since then, the data has continued to move in the same direction. Year-to-date pro forma volume of $189 billion through early July is up more than 20% from the $157 billion priced at the same point last year. In Europe, high yield BB trailing twelve-month issuance is hovering near its all-time highs reached earlier this year. And the composition of that issuance tells the more interesting story. BB-rated paper has swung to a 59% share of month-to-date June supply, up sharply from 30% in May, while single-B issuance has collapsed to less than 9%, a 15-month low.
Transactions that would previously have been financed entirely in the loan market are increasingly requiring bond market participation to be completed, with EA Sports and Sealed Air being prime recent examples: the loan market could not absorb the transaction alone, and bond market participation was what got the deal done. We do not believe that outcome is an isolated episode, but rather a signal of something structural. In fact, ~7% of the loan market matures within two years, which means >$100bn of par value in the loan market will need to consider if high yield is a better option. This is in addition to the ~10% of the high yield market that matures within the next two years, the highest share since the pre-GFC period.
While new transactions have increasingly leaned on the high yield market, we believe upcoming refinancings are likely to include high yield in a more significant way. Asurion offers an instructive example. As recently as last year, Asurion was a loan-only issuer with approximately $12 billion outstanding across half a dozen tranches. Facing a near-dated maturity profile spanning 2026 through 2028, and with many existing investors already at capacity on the name, the company needed a new approach. By pivoting to high yield, it reached a differentiated investor base and priced $3.3 billion of secured bonds to repay existing tranches earlier this year. The paydown restored confidence in the loan market, drove those remaining tranches tighter, and made subsequent refinancings cheaper. The virtuous cycle that followed was not accidental, but rather a direct consequence of opening a new financing channel. Looking ahead, we expect many software issuers, particularly those demonstrating strong cash generation and consistent profitability, to follow a similar path. The high yield market offers not just an alternative, but in some cases, the better route.
What does this mean for investors?
The case for high yield today is not a bet on spread compression, but rather a structural reallocation thesis with multiple return paths. For allocators building exposure across both markets, the U.S. and European opportunity sets also don't move in lockstep — sector composition, issuer base, and rate-cycle timing differ enough that combining them adds a diversification benefit distinct from either market alone.
The first is the yield itself. With the ICE BoA High Yield index at roughly ~7.3% today, the asset class offers attractive absolute income at a moment when fixed-rate certainty is increasingly valuable. In Europe, ICE BoA European High Yield a current yield of ~5.5%, with FX hedged investors picking up an additional ~150 basis points of carry. A bond that locks in that coupon for five to eight years, with call protection, is a different proposition from a floating-rate loan that reprices every quarter in a volatile rate environment.
Notably, U.S. Morningstar LSTA leveraged loan index yields currently sit at 8.1%, ~100 basis points above the high yield bond index. For investors willing to move up in quality, that premium no longer reflects better fundamentals. It reflects software overhang, CLO technical pressure, and documentation concerns that have accumulated in the loan market. In other words, the loan market is offering more yield today precisely because it carries more risk. High yield, at current levels, may be the more compelling risk-adjusted entry point.
The second is the optionality embedded in the structure. When issuers refinance early, as they have done consistently in periods of improving credit access, investors who purchased at a discount capture returns well above the modeled yield to worst. When bonds are not called and remain outstanding above par, the coupon continues to compound, often delivering realized returns that exceed initial expectations. When M&A activity accelerates, bonds held at a discount get called at par, generating IRRs that spread tightening alone would never produce.
For allocators who have underweighted high yield or have not revisited the thesis in several years, the question is simple: does the portfolio reflect what this market has actually become, or what it used to be? Structural quality is improving, issuer demand is growing, and the absolute return relative per unit of risk taken is a compelling proposition. The first act was a decade in the making. The second is already underway.
An op-ed by Christopher Sheldon, published August 3, 2026 in the Financial Times
预览 PDF
正在载入文档……