London-based KBC alpha Asset management is the specialist Japan and Asia regional focused fund of funds division of KBC Alternative Investment Management, an arm of KBC Bank & Insurance Holding NV, one of Belgium's leading financial institutions. Neale Safaty, KBC alpha's CIO, has over twenty years experience in the Japanese and Asian capital markets.

It is only quite recently that it was possible to construct a diversified portfolio of hedge funds that generate absolute returns predominantly from Japan. In fact, approximately 80% of the universe of Japan-focused hedge funds has emerged only in the last four years. Whilst a few Japan-only, equity long/short funds did exist several years ago, the bulk of hedge fund exposure to Japan was traditionally found in the Japanese allocations of global CB arbitrage funds and in the Japan portfolios of global macro funds. Index and warrant arbitrage strategies (which were prevalent in the late 1980's and early 1990's, when an "arbitrage" still existed) were employed primarily by the proprietary desks of investment banks and when run by hedge fund managers, were not offered as stand-alone products.

As of March 2004, we have identified 203 funds that are Japan-only or have a significant exposure to Japan in Pan Asian funds. This universe includes strategies from across the hedge fund risk/reward spectrum: opportunistic, event-driven and relative value. However, the lion's share of the investible universe is comprised of equity long/short and our Japan portfolio reflects this.

The remit of this column is to explain our investment criteria. Fund selection is a function of our overall portfolio objectives and risk management guidelines as well as the underlying individual fund appeal, although there are other factors that may be taken into consideration. At the macro level, the mandate of our fund allows us to invest across strategies. Let us consider some of the key ones.

As stated, the universe of Japan-focused hedge funds is dominated by equity long/short and to a great extent this will inevitably heavily influence the weightings of a portfolio, unless an allocator feels very strongly that Japan long/short cannot generate good risk-adjusted returns.

We would argue that this is unlikely for several reasons. Firstly, the size, depth, breadth and maturity of Japan's capital markets and derivatives capability means that it is cheap and easy to trade from both a long and a short perspective. Secondly, despite the improvement in the economy, the corporate profit recovery that is underway and the favourable policy initiatives that have been introduced by the authorities, the market can trade on factors that are not wholly fundamental. This disparity can offer tremendous opportunities that smart long/short managers can exploit.

Thirdly, the acceleration in corporate restructuring, including mergers and bankruptcies provide additional sources of alpha for long/short funds. Fourthly, despite the rapid growth in recent years of long/short funds, they do not represent a significant proportion of stock market activity and are still tiny, relative to the long-only equity universe, in terms of both assets under management and number of funds. Long/short is essentially operating in a hedge fund environment that is not crowded and managers tend not to 'chase' the same trades.

Finally, the huge surge in stock market volume that has been fuelled by foreigners purchasing equities as they discount improving fundamentals and a more favourable demand/supply outlook is a positive factor for long/short. For example, managers that discover opportunities in smaller cap stocks can be more aggressive when liquidity is so good and furthermore, execution impact is negligible when volumes are so high.

As managers with long-standing professional experience of Japan's convertible bond markets and with expertise in convertible bond arbitrage, we do not feel that it is appropriate, at present, to invest in convertible bond arbitrage for a number of reasons.

Firstly, the Japanese CB universe has contracted, as new financings have not kept pace with bonds have matured. Secondly, by historic standards, Japanese CBs are relative expensive. Thirdly, the overwhelming majority of convertibles are held by CB arbitrage funds. Whilst long/short funds do invest in CBs when appropriate, equity funds' ability to employ them as stock substitutes is limited as the vast majority are deep-out-of-the money. Furthermore, at new issue, the pricing of CB's does not attract equity-substitute interest. We regard an environment in which an asset class is predominantly held by one investor-type as an unhealthy one. Should any of these factors change, we would incorporate CB arbitrage.

Equity statistical arbitrage is a strategy that conceptually interests us and we have one manager operating in this space. However, we believe that in today's environment, our portfolio is better positioned with a heavy bias towards long/short.

Event-driven strategies have relatively low representation in the Japanese hedge fund universe. We have taken a strategic view to exclude real estate/distressed securities funds from our portfolio at present . As the underlying assets are illiquid, the funds may require long work-out periods to generate decent returns and thus tending to have lock-up clauses.

However, as far as the risk arbitrage/corporate activity space is concerned, we are keen to increase our allocation. Two of our more recent investments have been in variants of this strategy. We believe that changes to the legal, regulatory and cultural framework will lead to significant and increased corporate activity in Japan. We have already witnessed diverse examples of this is recent years, such as, less emphasis placed on protecting corporate stakeholders, share buy-backs, consolidated accounting, corporate restructuring (like equity for debt swaps), easier merger procedures and spin-offs and the absorption of affiliates, to name but a few.

We are interested in Japan dedicated macro as a strategy diversifier although this is very limited at present. In the global hedge fund context, macro represents approximately 8% of the universe. However, as specialist regional macro is far more limited in scope, we would expect this to remain a small minority, but increasing, strategy.

From a portfolio perspective, the addition of a new underlying fund has to be consistent with the risk/return objectives of our fund and risk management guidelines. We are looking to generate, annualised returns in the region of 10-12% with a low volatility, around 3-5%. This is on the basis that the fund is not leveraged. We recognise that in certain periods it may be possible to generate much higher returns and of course, in difficult years, grinding out single-digit returns may, in fact, be a "good" performance.

We believe that a portfolio of between 12 and 20 funds is appropriately diversified for risk control, yet sufficiently concentrated to produce target returns. We currently own 16 funds. We are keen to ensure that our portfolio consists of funds that have relatively low correlation to each other although we are less concerned if funds have a degree of correlation to bull phases as long as they exhibit low correlation to the downside. A potential new fund is regarded as either a diversifier or as a substitute and a correlation overlay will be undertaken when determining a funds possible status. A new fund will also be labelled as a low-, medium- or a high-risk fund, based on a combination of the investment strategy and the entity risk. Our portfolio consists of funds across the risk spectrum, but is consistent with our overall target return and risk objectives.

Of those funds that are included on our short-list, we categorise them by our likelihood to invest. Tier one funds are those we are very keen on. Funds in tier two are those we are interested in but where we have issues or where we have not completed due diligence. In tier three are funds that we continue to monitor but at present are impressed with.

We impose limits to single positions, on top-five fund concentration and on redemption liquidity. We are not averse to investing in funds with lock-up periods or low liquidity, especially as it is sensible in certain strategies to impose these restrictions, although we tend to favour underlying funds that have monthly liquidity. We also prefer to invest with managers that conduct currency hedging at the fund level, as we are currency neutral. We will hedge ourselves, however, if they do not.

We are flexible on a number of issues. Whilst we have not seeded new managers, we are willing to invest on day one if the key return generators are previously or currently known to us in professional capacity. A specific fund track record is not a prerequisite for investment.

Our portfolio consists of institutional, boutique and quasi-institutional entities (managers that have significant assets, extensive infrastructure and possibly several funds in the stable but are not regarded as institutions per se). We recognise that there are advantages and disadvantages of both institutions and boutiques, although the inclusion of institutions reduces individual entity risk and therefore, the overall risk of the portfolio.

We are also pragmatic on the location of where a fund is managed and by whom. Our fund is comprised of Gaijin and Japanese managers and by entities operating in Japan and outside of Japan. We have analysed the performance of managers throughout our universe and there is no statistically significant evidence that risk-adjusted returns from funds based in Japan are superior to those funds operating elsewhere in the world, despite advantages that local presence and intuitiveness may bring.

We are open-minded on asset size. If we are analysing a boutique, we will consider the breakeven amount of assets and invariably do not invest until this point is reached. It may transpire that we will invest in a fund that has a relatively low asset base. It is also possible that we may commit to a fund that is small and in which we are initially a significant or relatively large investor. This has been the case where a boutique has a significant amount of assets that include Japan exposure, but where there is a small dedicated Japan-only product or when the entity is an institution but the fund is either young or lacking assets because of, say, low-profile marketing.

The universe of Japan hedge funds is dynamic. Our Japan fund was launched in January 2002 and since inception, the number of funds has expanded quite rapidly, as we envisaged. In fact, it is very encouraging that there are an increasing number of very talented managers emerging from long-only institutions and investment banks.

Originally, we clipped the wings of the universe that we identified by excluding funds where assets were unjustifiably low relative to the age of the fund, where performance was poor or where key managers were personally known to us and we felt that they were not strong enough. Investing in hedge funds is a people business; we do not trade them like stocks and thus the integrity of managers is very important.

Notwithstanding, there are now funds brought into the fold that we had initially excluded from our monitor list because circumstances changed. It is also a matter of procedure to meet all start-ups unless we know the managers and it is highly unlikely that we will not invest. We insist upon at least one site visit before committing to a fund and we will conduct more extensive due diligence on boutique managers.

It is not uncommon for fund of funds to monitor investment candidates for a specific period. We do not set a minimum time-frame, evident by the fact that we have, on occasion, invested on day one. However, it is not unprecedented for us to follow a manager or a fund for in excess of twelve months before pulling the trigger.

Ideally, once committed, we like to make investments with an eighteen-month time-horizon in mind. We believe that over the course of about eighteen months, it is likely that there will be at least one period when the environment is benign for a particular strategy/style and there will probably be at least one period where conditions will be difficult. Furthermore, we ought to be able to assess whether our manager is achieving risk-adjusted returns that are within or above our expectations. This time-frame should be a sufficiently long enough period to confirm whether our manager can make decent returns while the "sun is shining" and to determine whether investors capital can be preserved when the going is tough, i.e. to get a sense of whether they are good hedge fund managers or not. Of course, it is not given that we will be invested for a minimum of eighteen months. Our post-investment review criteria includes, amongst other things, peer group analysis, possible strategy or style drift, infrastructure or personnel changes, asset contraction or rapid growth and events that may specifically impact expected strategy/style returns, such as short-selling restriction or lack of corporate activity.

So, what do we consider when we analyse a specific manager? In our preliminary meeting or meetings with prospective candidates, we will focus on the fund's historic returns, the investment process, risk management, the key portfolio managers and/or analysts and an overview of the organisational infrastructure and fund structure. We give priority to qualitative research rather than quantitative research.

There are several aspects to our qualitative research. Fundamental to assessing any fund is to understand and to evaluate fully the investment process and approach. Our strength is that the key members of our team have extensive experience in the Japanese capital markets and are able to discuss managers' investment approach to Japan against a back-drop of our deep personal knowledge of the securities, companies and market place. We are also able to review performance in different market conditions with the advantage that we know when an environment enables managers to just "pick up dollars lying on the floor" or whether they really had to earn their salt.

Risk management is very important to us, we like to discuss with managers how they would expect to cope under different types of conditions so that, should we invest, we minimise negative surprises. Furthermore, we like to discuss the rationale for their specific risk management policy and their enforcement procedures. We regard changes in risk management practice positively as we believe that it is likely to evolve. Even so, we do not necessarily take issue if risk management policy is the same as it was on day one. It is interesting to note that we have invested in funds that had poor first years but have demonstrated to us, that risk management has been enhanced.

We are fortunate that we have an extensive personal knowledge of many of the Japan-focused hedge fund players and, therefore, we like to understand why they have chosen this path and why and how they developed their specific business models. We believe that we can determine their competitive advantages and areas of weakness and we will examine potential issues that may impact their strategy and their ability to generate absolute returns. If they are boutiques, we like to identify growth plans for the management company and clarify their remuneration culture and ownership structure. We will also assess the possible capacity to us as potential investors as we would like our investment to grow as our business expands.

Whilst we emphasis qualitative research, we naturally conduct quantitative research. We review risk, return and correlation data, and we make peer group comparisons. We believe that there are limitations in this analysis and quite often dismiss results that may, on the face of it, be regarded as negative or even positive, where we feel the data is not statistically significant.

We tend to focus on cross-correlation analysis with qualitative overrides. For example, if existing Fund A appears to be highly correlated with a potential fund, Fund Y, if the strategy, style or investment approach differs significantly, we would not be too concerned. If, however, we are analysing two long-short funds that have operated over a similar statistically significant period with, say, a large cap top 400 stock mandate and where the managers have emerged from the same stable, the high correlation would have more emphasis in our selection process.

Once we are comfortable that we like the process, risk management and people and that we sense a manager may be either a diversifier or a replacement fund for us, we will conduct our operational due diligence, legal review and reference checking. In our operational due diligence, we review, for example, systems, trading platforms, disaster recovery plans, choice of service providers, compliance procedures and the fund valuation process and degree of transparency, to name but just a few. The legal review will primarily focus on the fund's prospectus although, as and when appropriate, we consider other documentation. One of our advantages in assessing the character, integrity and strengths of managers is that we are able to draw upon our decades of experience and professional network. If we have not had personal dealings with key principals in the past, we are able to solicit opinion from our colleagues within the KBC Group and we also are able to leverage the network of our professional acquaintances on both the buy and sell side of the business.

To conclude this column, it is worth making some general comments on management selection issues that have arisen during the course of researching the universe and since the fund's inception.

We tend to prefer investing in funds where the managers have worked together in some capacity at other organisations. We are not generally keen on hedge funds that are managed by ex-brokers or a team that may lack balance, i.e. where the key return drivers are analysts and portfolio managers that have no obvious trading acumen. Conversely, with respect to non-relative value strategies, we are not enthusiastic about funds run by ex proprietary traders. We also prefer the key functions, trading, research/portfolio construction, operations and compliance to be separated, if possible. Nevertheless, we strive to be pragmatic and open-minded. An experienced and talented equity long-only manager from an institutional background may not be a good equity long-short manager, whereas a bank's proprietary CB trader may be in fact, an excellent equity long-short manager.

Running a successful fund of funds portfolio is about finding the right people and putting them together in the right shapes and sizes. It is, certainly, more art than science.